Information

How to do western analysis to lung tissue?

How to do western analysis to lung tissue?


We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

I am working with lungs from mice, and I want to do a western blot analysis to my samples. I am having trouble with this because my images turn out very "messy", with a high background. Perhaps I should filter my samples following their homogenization? Has anyone had this problem and knows how to avoid it? What is the best method to homogenize the samples?


Doing westerns with primary tissue can be tough, especially because of the presence (sometimes at high levels) of extracellular matrix (ECM). This material can be quite resistant to homogenization and some lysis buffers. One method I have found to work is to snap-freeze the tissue in liquid nitrogen (not on dry ice or at -80°C) directly after removal from the animal, or after any sub-dissection you may wish to do. Then, when you are ready to prepare your sample, grind it thoroughly with a mortar and pestle in LN2, collect the dust and add a good lysis buffer like RIPA at 2-3X concentration, along with protease and phosphatase inhibitors, also at 2-3X. Then, instead of simply homogenizing, use a probe-tip sonicator (not a water bath one) and sonicate the lysate on ice with 8-10 pulses of perhaps 3-5 seconds each, ensuring that the lysate does not heat up too much, and making sure it does not froth up by getting the tip too close to the surface. Sonication sends very powerful shearing forces through liquid and is excellent at destroying membranes and breaking apart ECM. From here you can do your protein concentration assay (remember to use the same concentration of lysis buffer as your blank), aliquot the lysate and store at -80, or add SDS sample buffer and run your gel.

I routinely use sonication on all of my western samples, regardless of whether it's tissue or cell lines I'm dealing with, and I almost never get smeared lanes anymore. Sonication reduces the viscosity or "thickness" of the lysate, allowing you to pipette it much more easily and accurately. It is also useful when your sample "coagulates" after adding SDS sample buffer, and in this instance it doesn't matter if the sample gets hot, as you'll just be boiling it anyway.

I hope this helps, please let me know if you're still having problems. I used to do troubleshooting like this for a living :)


An integrative pharmacogenomics analysis identifies therapeutic targets in KRAS-mutant lung cancer

KRAS mutations are the most frequent oncogenic aberration in lung adenocarcinoma. KRAS mutant isoforms differentially shape tumour biology and influence drug responses. This heterogeneity challenges the development of effective therapies for patients with KRAS-driven non-small cell lung cancer (NSCLC).

Methods

We developed an integrative pharmacogenomics analysis to identify potential drug targets to overcome MEK/ERK inhibitor resistance in lung cancer cell lines with KRAS(G12C) mutation (n =�). We validated our predictive in silico results with in vitro models using gene knockdown, pharmacological target inhibition and reporter assays.

Findings

Our computational analysis identifies casein kinase 2A1 (CSNK2A1) as a mediator of MEK/ERK inhibitor resistance in KRAS(G12C) mutant lung cancer cells. CSNK2A1 knockdown reduces cell proliferation, inhibits Wnt/β-catenin signalling and increases the anti-proliferative effect of MEK inhibition selectively in KRAS(G12C) mutant lung cancer cells. The specific CK2-inhibitor silmitasertib phenocopies the CSNK2A1 knockdown effect and sensitizes KRAS(G12C) mutant cells to MEK inhibition.

Interpretation

Our study supports the importance of accurate patient stratification and rational drug combinations to gain benefit from MEK inhibition in patients with KRAS mutant NSCLC. We develop a genotype-based strategy that identifies CK2 as a promising co-target in KRAS(G12C) mutant NSCLC by using available pharmacogenomics gene expression datasets. This approach is applicable to other oncogene driven cancers.

This work was supported by grants from the National Natural Science Foundation of China, the National Key Research and Development Program of China, the Lung Cancer Research Foundation and a Mildred-Scheel postdoctoral fellowship from the German Cancer Aid Foundation.


Centrifugation removes lipids and particulate matter, such as cell debris. If the sample is still not clear after centrifugation, use filter paper or a 5 μm filter as a first step and one of the filters below as a second step filter.

  • For small sample volumes or proteins that adsorb to filters, centrifuge at 10 000 × g for 15 min.
  • For cell lysates, centrifuge at 40 000 to 50 000 × g for 30 min.
  • Serum samples can be filtered through glass wool after centrifugation to remove any remaining lipids.

Flow Cytometric Analysis of Macrophages and Dendritic Cell Subsets in the Mouse Lung

The lung hosts multiple populations of macrophages and dendritic cells, which play a crucial role in lung pathology. The accurate identification and enumeration of these subsets are essential for understanding their role in lung pathology. Flow cytometry is a mainstream tool for studying the immune system. However, a systematic flow cytometric approach to identify subsets of macrophages and dendritic cells (DCs) accurately and consistently in the normal mouse lung has not been described. Here we developed a panel of surface markers and an analysis strategy that accurately identify all known populations of macrophages and DCs, and their precursors in the lung during steady-state conditions and bleomycin-induced injury. Using this panel, we assessed the polarization of lung macrophages during the course of bleomycin-induced lung injury. Alveolar macrophages expressed markers of alternatively activated macrophages during both acute and fibrotic phases of bleomycin-induced lung injury, whereas markers of classically activated macrophages were expressed only during the acute phase. Taken together, these data suggest that this flow cytometric panel is very helpful in identifying macrophage and DC populations and their state of activation in normal, injured, and fibrotic lungs.

Flow cytometry is a mainstream tool for studying the immune system. However, a systematic flow cytometric approach to identify subsets of macrophages and dendritic cells (DCs) accurately and consistently in the normal mouse lung has not been described. Here we developed a panel of surface markers and an analysis strategy that accurately identifies all known populations of macrophages and DCs, and their precursors in the lung, during steady-state conditions and bleomycin-induced injury.

Cells of the innate immune system, and especially myeloid cells such as neutrophils, eosinophils, monocytes, macrophages (alveolar and interstitial), and dendritic cells (DCs, i.e., plasmacytoid DCs, CD103 + DCs, and CD11b + DCs), play an important role in lung development and physiology, and contribute to important lung diseases, including pulmonary infection, cancer, asthma, chronic obstructive pulmonary disease, and pulmonary fibrosis (1–5). Alveolar and interstitial lung macrophages exhibit different origins and life spans in lungs, and have been identified as key regulators of pathological and reparative processes. Alveolar macrophages, which are considered tissue-resident macrophages, populate lung tissue during early embryogenesis and remain viable for prolonged periods, with minimal replenishment from bone marrow–derived monocytes (6). In contrast, interstitial macrophages originate from bone marrow–derived monocytes and have a shorter half-life (7, 8). In recent studies, several groups of investigators suggested that these two populations of lung macrophages play opposing roles in lung injury. Alveolar macrophages appear to limit neutrophil influx into the lung during acute lung injury (9) or chronic exposure to organic dust (10), whereas interstitial macrophages promote neutrophil extravasation (11, 12). An additional layer of complexity is added by the phenotypic plasticity of macrophages. Classically activated macrophages (sometimes referred to as M1-polarized) have been suggested to promote the development of acute lung injury, whereas alternatively activated macrophages (M2) may play a role in limiting or resolving lung inflammation (13) and/or potentially promoting the development of fibrosis (14–17).

Understanding the roles played by these different macrophage populations and macrophage phenotypes in the pathophysiology of lung injury, repair, and fibrosis requires proper identification, enumeration, and phenotypic characterization. Flow cytometry has become an essential tool for analyses of the immune system, because it offers a short turnaround time between sample preparation, acquisition, and analysis, allows for the accurate enumeration of individual cell subsets (including very rare subsets), and provides an opportunity for detailed molecular phenotyping. However, flow cytometric analyses of innate immune cells are challenging even in the normal lung, and these problems are magnified in the presence of lung inflammation or fibrosis.

Recently, Gautier and colleagues used gene-expression profiling to identify differentially expressed genes in tissue macrophages, compared with other tissue-resident myeloid cells (18). We combined some of these newly identified markers with those described elsewhere in the literature to develop a panel of antibodies for use in flow cytometry that meets three a priori criteria, namely, the panel (1) clearly distinguishes between different myeloid populations in the mouse lung, (2) relies exclusively on surface markers to allow for live cell sorting, and (3) performs well in the injured or fibrotic lung. We used this panel to identify a minimal set of surface markers that can be used by investigators without access to eight or more parameter instruments. We validated the usefulness of both full and minimal panels during early acute lung injury and the late fibrosis after intratracheal administrations of bleomycin, and we evaluated the utility of these markers for differentiating between what are currently understood as M1 and M2 macrophage subsets. This systematic approach to flow cytometric analyses of the innate immune system in the mouse lung should prove helpful in improving our understanding of the role that individual subsets of macrophages and DCs play in the development of lung disease.

Eight-week-old male C57BL/6 mice were purchased from Jackson Laboratory (Bar Harbor, ME) and housed at a barrier-free and specific pathogen–free facility at the Center for Comparative Medicine at Northwestern University (Chicago, IL). All procedures were approved by the Institutional Animal Care and Use Committee at Northwestern University.

Mice were anesthetized with isoflurane, their lungs were intubated orally with a 20-gauge Angiocath (Franklin Lanes, NJ), and two 50-μl sterile aliquots of PBS (control) or 0.025 IU bleomycin (AAP Pharmaceuticals LLC, Shamburg, IL) were instilled through the catheter, 3 minutes apart. After each aliquot, the mice were placed on their right side and then left side for 10–15 seconds. Mice were killed 5 and 21 days after the instillation of bleomycin.

After the mice wee killed, their lungs were perfused through the right ventricle with 5 ml of PBS. The lungs were removed, and the large airways were dissected from the peripheral lung tissue. The peripheral lung tissue was cut into small pieces with scissors, transferred into C-tubes (Miltenyi, Auburn, CA), and processed in digestion buffer (1 mg/ml of Collagenase D and 0.1 mg/ml DNase I, both from Roche, Indianapolis, IN, in Hanks’ balanced salt solution) and a GentleMACS dissociator (Miltenyi), according to the manufacturer’s instructions. Homogenized lungs were passed through 40-μm nylon mesh to obtain a single-cell suspension. The remaining red blood cells were lysed using BD Pharm Lyse (BD Biosciences, San Jose, CA). The resultant cells were counted using a Countess cell counter (Invitrogen, Carlsbad, CA). Dead cells were excluded, using trypan blue.

Cells were stained with viability dye Aqua (Invitrogen) or eFluor 506 (eBioscience, San Diego, CA), incubated with FcBlock (BD Biosciences), and stained with a mixture of fluorochrome-conjugated antibodies (Table for a list of antibodies, clones, fluorochromes, and manufacturers). Data were acquired on a BD LSR II flow cytometer using BD FACSDiva software (BD Biosciences see Table E2 for instrument configuration), and compensation and data analyses were performed “offline” using FlowJo software (TreeStar, Ashland, OR). Cell sorting was performed on a FACSAria II instrument (BD Biosciences) with the same configuration as the LSR II. Cytospins were prepared from sorted cells and stained with a Diff-Quik Stain Set (Siemens Healthcare, Malvern. PA). “Fluorescence minus one” controls were used when necessary. Cell populations were identified using sequential gating strategy (see R esults ), and the percentage of cells in the live/singlets gate was multiplied by the number of live cells (after trypan blue exclusion) to obtain an absolute live-cell count. The expression of activation markers is presented as median fluorescence intensity (MFI).

Differences between groups were determined according to ANOVA. When ANOVA indicated a significant difference, individual differences were examined using t tests with a Tukey correction for multiple comparisons, as indicated. All analyses were performed using GraphPad Prism, version 5.00 (GraphPad Software, San Diego CA). Data are shown as means ± SEMs. P < 0.05 was considered statistically significant in all tests.

To identify myeloid populations in lungs accurately under steady-state conditions, we performed 10-color flow cytometry and sequential gating analysis ( Figure 1 ). After the exclusion of doublets and debris, immune cells were identified using the pan-hematopoietic marker CD45. Dead cells may also be excluded at this step, using live/dead staining (Figure E1). In normal mouse lungs, alveolar macrophages were readily identified, based on the expression of Siglec F, CD11c, CD64, F4/80, the absence of CD11b, high side scatter, and high autofluorescence (Table 1, Figure 1A , and Figure E2). CD103 + DCs were enumerated based on their high expression of CD11c, CD24, CD103, and major histocompatibility class II (MHC II), and their absence of CD11b. Neutrophils express Ly6G, which is not detected in any other cell types, thus allowing for their clear and unambiguous identification. Eosinophils were identified based on their expression of Siglec F and F4/80, and after gating out alveolar macrophages (which also express high concentrations of Siglec F). High concentrations of CD11b and CD24, as well as high side scatter and the absence of CD11c and MHC II molecules, provided additional help in the identification of neutrophils and eosinophils. Moreover, the high expression of CD11b permitted the separation of monocytes, interstitial macrophages, and CD11b + DCs from natural killer cells, which expressed intermediate concentrations of CD11b. Interstitial macrophages and CD11b + DCs expressed MHC II and high-to-intermediate concentrations of CD11c, and were distinguished based on their expression of CD64 and CD24, correspondingly. CD11b hi cells do not express MHC II and express only low concentrations of CD64, and were separated into Ly6C + and Ly6C − subsets. The expression of CD64, MHC II, and activation markers such as CD40, CD80, and CD86 was low in Ly6C + and Ly6C − subsets (data not shown). Plasmacytoid DCs were identified as mPDCA-1 + , as well as CD11c int B220 + (Figure E3).

Figure 1. Gating strategy used to identify myeloid-cell subsets in the normal mouse lung. Cells were isolated from enzymatically digested mouse lungs, and after the exclusion of doublets and debris, immune cells were identified by CD45 staining. (A) A sequential gating strategy was first used to identify populations expressing specific markers: alveolar macrophages (MΦ) (Siglec F + CD11b − CD11c + CD64 + ), CD103 + dendritic cells (DCs) (CD11c + CD103 + CD24 + ), neutrophils (CD11b + Ly6G + ), and eosinophils (Siglec F + CD11b + CD11c − ), followed by the identification of populations with overlapping expression patterns: interstitial macrophages (CD11b + MHC II + CD11c + CD64 + CD24 − ), CD11b + DCs (CD11b + MHC II + CD11c + CD24 + CD64 − ), and monocytes (Mo)/undifferentiated macrophages (CD11b + MHC II − CD64 +/− Ly6C lo ). Scale bar in microphotographs = 5 μm. (B) Identification of macrophages and DCs using the minimal panel of surface markers. Both alveolar macrophages and CD103 + DCs are identified as CD11b − CD11c + cells, and are further separated using CD64 and CD24, correspondingly. If necessary, MHC II can be used to confirm gating in CD103 + DCs (not shown). Gating on CD11b hi cells allows for the separation of myeloid cells from lymphoid cells that either do not express this marker (T and B cells), or express it at intermediate level (natural killer cells). Granulocytes (neutrophils and eosinophils) can be gated out as CD24 + CD11c − , and the identification of CD11b + DCs (CD11b + MHC II + CD11c + CD24 − CD64 − ), interstitial macrophages (CD11b + MHC II + CD11c + CD64 + CD24 − ), and monocytes/undifferentiated macrophages can be continued as in the full panel (CD11b + MHC II − CD64 +/− Ly6C lo ). FSC, forward scatter MHC II, major histocompatibility complex class II SSC, side scatter.

TABLE 1. PHENOTYPES OF MYELOID CELLS IN THE NORMAL MOUSE LUNG

Definition of abbreviations: AM, alveolar macrophages DCs, dendritic cells IM, interstitial macrophages Mo, monocytes MΦ, alveolar macrophages MHC II, major histocompatibility complex class II NT, not tested.

Symbols indicate the expression of a given marker. +, high expression ±, low or intermediate expression −, absence of expression. For a more differentiated assessment of expression, see Figure E2 in the online supplement. The minimal set of surface markers required for the accurate identification of macrophages and DCs in the mouse lung is shown in boldface.

To validate the results of flow cytometric experiments, we sorted individual populations from normal mouse lungs and examined their morphology ( Figure 1A ). CD103 + and CD11b + DCs and interstitial macrophages exhibited irregularly shaped nuclei and numerous vacuoles in the cytoplasm, consistent with their role as phagocytes. Alveolar macrophages exhibited a similar morphology, but with more prominent pseudopodia. Ly6C + and Ly6C − populations contained both monocyte-like and macrophage-like cells. The monocyte-like cells contained bean-shaped nuclei without vacuoles, and the macrophage-like cells contained irregularly shaped nuclei with numerous vacuoles.

Although our approach allows for the accurate identification of all myeloid subsets in the mouse lung, it requires instruments capable of analyzing at least 10 fluorescent parameters, and these instruments may not be available to all investigators. Therefore, we developed a minimal set of surface markers and corresponding antibodies that allowed for a clear identification of all populations of myeloid cells in normal mouse lungs ( Figure 1B ). This approach identified well-defined populations such as alveolar macrophages and CD103 + DCs, while allowing for the separation of monocyte-derived populations such as interstitial macrophages (CD11b + CD11c + CD64 high MHC II + CD24 − ), CD11b + DCs (CD11b + CD11c + CD64 − MHC II + CD24 + ), and less mature monocytes and macrophages (CD11b + CD11c +/− CD64 low MHC II − CD24 low ).

We examined our panel of markers for the identification of myeloid populations in the lung in the bleomycin model of lung injury followed by fibrosis. Previous studies have shown that an intratracheal administration of bleomycin results in acute lung injury, which becomes maximal 3 to 5 days after instillation and subsequently resolves. This is followed by the transforming growth factor–β–mediated development of lung fibrosis, which peaks between Days 21 and 28 after injury, and resolves slowly thereafter (16). In the acute (Day 5) and fibrotic (Day 21) phases of bleomycin-induced lung injury, the full and minimal panels of surface markers readily identified all myeloid subsets in homogenized lung tissue ( Figure 2 ). During the acute phase (Day 5), the number of alveolar macrophages significantly decreased, whereas the number of interstitial macrophages and CD11b + DCs increased ( Figure 3A ). In contrast, during the fibrotic phase (Day 21), the number of alveolar macrophages in bleomycin-treated animals was higher than in control animals, whereas the number of interstitial macrophages returned to control concentrations ( Figure 3A ). An increase in the number of alveolar macrophages during the fibrotic phase coincided with the emergence of a new subpopulation of Siglec F low alveolar macrophages. These Siglec F low alveolar macrophages expressed CD11b and elevated concentrations of CD11c, CD14, CD36, and CD64 (Figure E4). Siglec F was required for their identification, because neither CD11b nor CD11c allowed for the clear separation of these two subsets of alveolar macrophages. Importantly, even 38 days after bleomycin-induced lung injury, the Siglec F low population was still present, and the ratio of Siglec F high to Siglec F low macrophages was unchanged (data not shown).

Figure 2. The phenotype of myeloid cells in mouse lungs changes during the course of bleomycin-induced lung injury. Left to right: Normal lung, 5 and 21 days after instillation of bleomycin. Top images were gated on CD45 + cells, with neutrophils and eosinophils gated out.

Figure 3. Changes of myeloid-cell subsets in mouse lungs during bleomycin-induced lung injury (Days 5 and 21). (A) Numerical changes of myeloid-cell subsets were identified as described in Figure 1 . Values represent means ± SEMs. Differences between groups were compared using one-way ANOVA. ***P < 0.001. (B) Expression of markers associated with classically (CD40, CD80, and CD86) and alternatively (CD71, CD206, and RELMα) activated macrophages on alveolar and interstitial macrophages during bleomycin-induced lung injury. Values represent means ± SEMs for median fluorescence intensity (MFI) for the given marker. Differences between groups were compared using one-way ANOVA. **P < 0.01. ***P < 0.001.

During the acute phase of bleomycin-induced lung injury, the expression of CD64 and markers of “classically activated” or “M1-like” macrophages (CD40, CD80, and CD86) (19) were increased in both alveolar and interstitial macrophages ( Figures 3B and E5). However, during the fibrotic phase (Day 21), the expression of CD40 and CD80 returned to control concentrations ( Figure 3B ), whereas CD86 expression in alveolar macrophages remained elevated. In contrast to activation markers, the expression of CD71 (transferrin receptor), CD206 (mannose receptor), and resistin-like molecule alpha (RELMα), which are all associated with “alternatively activated,” “regulatory,” or “M2-like” macrophages (13, 19, 20), was elevated in alveolar macrophages during both the acute and fibrotic phases ( Figure 3B ). Importantly, no difference was evident in the expression of CD71, CD206, RELMα, and CD86 between Siglec F high and Siglec F low alveolar macrophages on Day 21.

Flow cytometric analyses of innate immune cells in the lung are complicated for several reasons. First, many myeloid cells, and particularly alveolar macrophages, exhibit high autofluorescence, which often decreases the resolution between “positive” and “negative” populations, leading to false positivity for a given antigen/fluorochrome (21–23). Moreover, the autofluorescence of myeloid cells in the lung may further increase after exposure to environmental particulate matter. Second, many populations of myeloid cells, and especially macrophages and DCs, express similar surface markers, which makes an accurate identification of individual cellular subsets using just one or two surface markers almost impossible (18, 24–26). We describe a panel of surface markers that can be used to identify different myeloid populations unambiguously in the mouse lung, using flow cytometry. In a well-described model of bleomycin-induced lung injury followed by fibrosis, we found that this panel was able to distinguish different myeloid populations and assess their level of activation. Unlike other approaches (26), this panel relies exclusively on surface markers, and therefore can be used to sort live cells for use in subsequent studies.

The normal mouse lung contains multiple populations of macrophages and DCs. Lung macrophages consist of two distinct populations, namely, alveolar macrophages that represent long-lived tissue-resident macrophages, and short-lived monocyte-derived interstitial macrophages. Alveolar macrophages play an important role in maintaining lung homeostasis by removing pathogens and noxious particles without inducing inflammation or recruiting monocytes and neutrophils (3, 4, 27). In contrast, monocyte-derived interstitial macrophages are recruited to the lung from the circulation in response to acute lung injury, and are major contributors to the inflammatory response (3, 4, 27). DCs in the lung are represented by monocyte-derived CD11b + DCs, plasmacytoid DCs, and CD103 + DCs, which originate from distinct precursors and play a crucial role in the induction and suppression of innate immune responses (2, 28). A precise identification of the relative and absolute compositions, as well as the activation state, of inflammatory populations in the lung is required if we are to understand their role in disease pathogenesis. This identification has been complicated by the absence of a specific macrophage or DC marker (18, 24) and the expression of similar surface markers in macrophages and DCs in lungs. Although plasmacytoid DCs and CD103 + DCs are easily identified using mPDCA-1 and CD103 antibodies, respectively (24, 29), discriminating between interstitial macrophages and CD11b + DCs in the lung is less straightforward (28). Historically, CD11b + DCs in the lungs were identified as CD11b + CD11c + MHC II + . However, this population was recently found to be heterogeneous (30). CD64, also known as Fc-gamma-Receptor 1 (FcγR1), has been shown to be useful in discriminating macrophages from CD11b + DCs in the mouse gut and muscle (31, 32), and together with proto-oncogene tyrosine-protein kinase MER (MerTK) and CD14, is one of the most specific macrophage markers (18). We found that a combination of CD64 with CD24 and MHC II allowed not only for the separation of CD11b + DCs from interstitial macrophages, but also for their discrimination from monocytes and undifferentiated macrophages.

Alveolar macrophages are long-lived tissue-resident macrophages. They populate the lung during early embryogenesis, and are able to maintain themselves for months with minimal replenishment from bone marrow–derived cells (6, 33). Moreover, resident alveolar macrophages have been shown to persist after LPS-induced or influenza A–induced acute lung injury, when they participate in the resolution of inflammation by phagocytosing apoptotic neutrophils and recruiting monocyte-derived macrophages (33). However, if alveolar macrophages are depleted using clodronate-loaded liposomes or the administration of diphtheria toxin to CD11c–diphtheria toxin receptor mice, they can be restored by monocyte-derived interstitial macrophages (7). The instillation of bleomycin induces the apoptosis of alveolar macrophages (34–36), which are then reconstituted from bone marrow–derived cells. Siglec F, which is typically considered an eosinophil marker (37), is highly expressed in murine alveolar macrophages, and when used in combination with CD11c or CD64, provides the most accurate identification of alveolar macrophages in the mouse lung. We found that in the normal mouse lung and during the acute phase of bleomycin-induced lung injury, alveolar macrophages maintain their distinct phenotype (Siglec F high CD11c + CD64 + CD11b − ) and can be easily separated from interstitial macrophages and CD11b + DCs, using flow cytometry. However, during the fibrotic phase, a new subpopulation of Siglec F low alveolar macrophages appears. In comparison to Siglec F high alveolar macrophages, these Silgec F low alveolar macrophages express higher concentrations of CD11b, CD11c, CD64, CD14, and CD36, and likely represent monocyte-derived interstitial macrophages taking on intermediate phenotypes as they differentiate into alveolar macrophages. Therefore, although an accurate identification of macrophage and DC subsets in the mouse lung is possible without using Siglec F, we found this marker very useful in providing additional information on the origins of alveolar macrophages. In contrast, F4/80, which is often considered a “classic” macrophage marker, did not separate interstitial macrophages, monocytes/undifferentiated macrophages, eosinophils, and CD11b + DCs. The addition of F4/80 to the panel provided no additional discriminatory power. Another “classic” macrophage marker, CD68, was recently proposed for the identification of myeloid cell subsets in the mouse lung (26). However, because CD68 is an intracellular marker, its use requires cellular fixation and permeabilization, precluding use of the cells in subsequent experiments. Our approach relies exclusively on surface markers, and therefore can be used to sort live cells by means of FACS.

In the past, many groups reported the results of two-parameter approaches to identify myeloid cell subsets in the murine lung. For example, investigators have used CD11b versus CD11c plots to identify alveolar and interstitial macrophages and even DCs (7, 23, 38). Our data suggest that although this approach allows for a fairly accurate identification of alveolar macrophages, it does not permit discrimination between CD11b + DCs, interstitial macrophages, and immature monocytes/macrophages. Furthermore, unless neutrophils and eosinophils are explicitly gated out before examinations of CD11b versus CD11c staining, they would fall into the CD11b + CD11c − region and possibly be incorrectly identified as monocytes and macrophages. Other multiparameter panels for analyzing the myeloid compartment of the mouse lung failed to discriminate between interstitial macrophages, CD11b + DCs, and immature macrophages/monocytes, and often failed to identify eosinophils (25, 26, 38).

Although flow cytometry provides a wealth of information about cell phenotypes, information about anatomical localization is essentially lost during sample preparation. One valid approach to overcome this problem involves the in vivo labeling of cells in the intravascular compartment with an anti-CD45 antibody (38). Another commonly used approach involves comparing the populations present in bronchoalveolar lavage with those present after enzymatic digestion of the lung. The origin of cells recovered from lavage fluid is attributed to the alveolar space, whereas cells recovered from the digested lung are attributed to the interstitium. However, results using this approach should be cautiously interpreted, because even after multiple lavages, only a fraction of cells can be recovered from the alveolar space (26, 39). Future studies might use immunohistochemical or immunofluorescence techniques to provide better correlations between cell surface markers and anatomic localizations in the intact lung.

During the past several years, the importance of macrophage polarization during inflammation and fibrosis has been increasingly recognized. Markers associated with “classically activated” “M1-like” macrophages are up-regulated only during the acute phase of bleomycin-induced lung injury, whereas markers associated with “alternatively” or “regulatory” “M2-like” macrophages are increased in alveolar macrophages during both the acute and fibrotic phases. Our data suggest that in the model of bleomycin-induced lung injury followed by fibrosis and repair, an “M2-like” macrophage response begins very early, in parallel with an initial “M1-like” response, rather than after its cessation. Unlike Listeria monocytogenes–infected peritoneum (40) or infarcted myocardium (41), the acute-phase “M2-type” response in bleomycin-treated mouse lungs is driven not by recruited monocytes, but by resident tissue macrophages. Of interest, similar findings have been reported in the mouse gut, where tissue-resident macrophages exhibited an anti-inflammatory profile, both in the native state and during acute inflammation (42). However, macrophage polarization is not limited to the M1 and M2 states, but also includes regulatory and resolution-phase macrophages (43, 44). Moreover, overlapping phenotypes and populations may exist simultaneously within the same tissue. Therefore, the proper assignment of macrophage polarization cannot be performed using only this limited number of surface markers. Perhaps analyses of gene expression in individually sorted populations of pulmonary myeloid-cell subsets during different stages of disease will allow for a better understanding of macrophage polarization status, and potentially help identify new markers and their associated functions.

In conclusion, we provide a flow cytometric approach to identify subsets of macrophages and DCs in the normal and inflamed mouse lung. Our panel can be used by investigators as a starting point to examine the role of resident and recruited macrophages and DCs in murine models of lung disease. When experimentally indicated, other markers, such as markers of neutrophils, plasmacytoid DCs, markers of macrophage activation, or viability dyes, can be included with this panel. This panel and its future refinements will provide a useful tool for investigators to examine the complex immune responses of the lung to its changing environment during health and disease.


The Lungs and Respiration

Air is supplied to the lungs through the process of breathing. The diaphragm plays a key role in breathing. The diaphragm is a muscular partition that separates the chest cavity from the abdominal cavity. When relaxed, the diaphragm is shaped like a dome. This shape limits space in the chest cavity. When the diaphragm contracts, it moves downward toward the abdominal area causing the chest cavity to expand. This lowers the air pressure in the lungs causing the air in the environment to be pulled into the lungs through air passages. This process is called inhalation.

As the diaphragm relaxes, space in the chest cavity is reduced forcing air out of the lungs. This is called exhalation. Regulation of breathing is a function of the autonomic nervous system. Breathing is controlled by a region of the brain called the medulla oblongata. Neurons in this brain region send signals to the diaphragm and the muscles between the ribs to regulate the contractions which initiate the breathing process.


Compartmentalization

Where in the tissue environment and by which cells a MMP is expressed and released are equally, if not more important, considerations in regulating the specificity of proteolysis than the affinity of the enzymesubstrate interaction. For example, in a test tube, matrilysin inactivates α1-antiproteinase inhibitor much more efficiently than does gelatinase-B. However, in tissue, at least in inflamed dermis, this serpin is selectively cleaved by neutrophil-derived gelatinase-B [22]. Matrilysin, an epithelial cell product, tends to be released lumenally away from the matrix (see later). An important concept is that cells do not indiscriminately release proteases. Rather, proteinases, such as MMPs, are secreted and anchored to the cell membrane, thereby targeting their catalytic activity to specific substrates within the pericellular space. Specific cell–MMP interactions have been reported in recent years, such as the binding of gelatinase-A to the integrin αvβ3 [33], binding of gelatinase-B to CD44 [28], and binding of matrilysin to surface proteoglycans [34]. Pro-gelatinase-A also interacts with tissue inhibitor of metalloproteinases (TIMP)-2 and MT1-MMP on the cell surface, and this trimeric complex is essential for activation of this gelatinase [35,36]. It is likely that other MMPs are also attached to cells via specific interaction to membrane proteins, and determining these anchors will lead to identifying activation mechanisms and pericellular substrates.

Cells also rely on surface receptors to 'sniff out' the presence and location of specific substrates. For matrix substrates, integrinligand contacts provide an unambiguous signal informing the cell of which protein it has encountered and, hence, which proteinase is needed and to where the enzyme should be delivered and released. A clear example of this type of spatial regulation is seen with collagenase-1 in human cutaneous wounds. Collagenase-1 is induced in basal epidermal cells (keratinocytes), in response to injury, as the cells move off the basement membrane and contact type I collagen in the underlying dermis [37]. Only basal keratinocytes in contact with dermal type I collagen express collagenase-1, and this inductive response is specifically controlled by the collagen-binding integrin α2β1, which also directs secretion of the enzyme to the points of cell–matrix contact [38]. This example demonstrates that expression and activity of a specific MMP can be confined to a specific location in an activated tissue (the superficial plane of a denuded epithelium) and to a specific stage of repair (re-epithelialization).


Pathologic Regulation of Collagen I by an Aberrant Protein Phosphatase 2A/Histone Deacetylase C4/MicroRNA-29 Signal Axis in Idiopathic Pulmonary Fibrosis Fibroblasts

Idiopathic pulmonary fibrosis (IPF) is characterized by the relentless expansion of fibroblasts depositing type I collagen within the alveolar wall and obliterating the alveolar airspace. MicroRNA (miR)-29 is a potent regulator of collagen expression. In IPF, miR-29 levels are low, whereas type I collagen expression is high. However, the mechanism for suppression of miR-29 and increased type I collagen expression in IPF remains unclear. Here we show that when IPF fibroblasts are seeded on polymerized type I collagen, miR-29c levels are suppressed and type I collagen expression is high. In contrast, miR-29c is high and type I collagen expression is low in control fibroblasts. We demonstrate that the mechanism for suppression of miR-29 during IPF fibroblast interaction with polymerized collagen involves inappropriately low protein phosphatase (PP) 2A function, leading to histone deacetylase (HDA) C4 phosphorylation and decreased nuclear translocation of HDAC4. We demonstrate that overexpression of HDAC4 in IPF fibroblasts restored miR-29c levels and decreased type I collagen expression, whereas knocking down HDAC4 in control fibroblasts suppressed miR-29c levels and increased type I collagen expression. Our data indicate that IPF fibroblast interaction with polymerized type I collagen results in an aberrant PP2A/HDAC4 axis, which suppresses miR-29, causing a pathologic increase in type I collagen expression.

Our work offers vital insight in type I collagen regulation in idiopathic pulmonary fibrosis fibroblasts and can potentially offer multiple therapeutic targets in the future.

Idiopathic pulmonary fibrosis (IPF) is a prevalent and progressive fibrotic lung disease that does not respond to therapy (1–5). IPF is characterized by a continuous expansion of the fibroblast population fibroblasts deposit type I collagen within the alveolar wall, causing scarred nonfunctional airspaces, progressive hypoxia, and death by asphyxiation (5–11). As the fibrosis evolves, there is a contiguous spread of the process from affected alveoli into adjacent anatomically normal gas exchange units, resulting in an uninterrupted reticular network of fibrotic tissue (12).

MicroRNA (miR)-29 is a potent regulator of type I collagen expression. Human miR-29 has three members—miR-29a, miR-29b, and miR-29c—that target COL1A1 and COL1A2 messenger RNA (mRNA) and decrease type I collagen expression (13). Prior work has determined that miR-29 levels are low in experimental models of pulmonary fibrosis and in IPF, whereas type I collagen expression is high (14, 15). This suggests an important role of miR-29 in regulating the excessive production of collagen in IPF. However, the mechanism leading to suppression of miR-29 in IPF remains unclear.

Our prior studies indicate that IPF lung fibroblasts manifest a distinct pathological phenotype characterized by an aberrant integrin signaling in response to interaction with polymerized type I collagen, a major component of the IPF fibrotic matrix (16–18). A seminal study by Heino and colleagues demonstrated that when normal fibroblasts interact with polymerized collagen, α2β1 integrin binds to collagen resulting in activation of protein phosphatase (PP2) A phosphatase (19). In contrast, when IPF fibroblasts interact with polymerized type I collagen, they fail to appropriately activate PP2A due to aberrant α2β1 integrin function, resulting in activation of proliferation signaling pathways (18).

Although the complete mechanism by which low PP2A activity promotes acquisition of the pathologic IPF fibroblast phenotype remains to be elucidated, recent studies have shown that PP2A regulates histone deacetylase (HDA) C4, a class II histone deacetylase whose function has been linked to TGF-β–mediated differentiation of normal fibroblasts (20). HDAC4 regulates gene expression by undergoing nucleocytoplasmic shuttling in response to environmental cues (21, 22). Class II HDACs contain an N-terminal regulatory domain that is subject to phosphorylation, and nucleocytoplasmic shuttling of HDAC4 is controlled by its phosphorylation state (21). Dephosphorylation by PP2A stabilizes HDAC4 and promotes its nuclear accumulation, whereas phosphorylated HDAC4 is located in the cytoplasm and is prone to degradation (21–24). Within the nucleus, HDAC4 represses transcription when tethered to a promoter. It does so by forming a complex with tissue-specific transcription factors, repressing their function and thereby regulating cell phenotype (22, 25–29).

A recent report has linked HDAC4 function to regulation of miR-29 (30). Because we have found that the α2β1 integrin/PP2A axis is abnormal in IPF fibroblasts during their interaction with polymerized type I collagen, we hypothesized that alterations in HDAC4 function may be responsible for suppression of miR-29 in IPF. Here we report that, in response to IPF fibroblast interaction with polymerized type I collagen, low PP2A function results in HDAC4 hyperphosphorylation and decreased HDAC4 nuclear import. We demonstrate that low levels of nuclear HDAC4 result in suppression of miR-29 and increased type I collagen expression. We have discovered that the mechanism involves HDAC4 regulation of miR-29. We have found that knockdown of HDAC4 suppresses miR-29 levels, whereas overexpression of HDAC4 increases miR-29 expression. Our data indicate that during IPF fibroblast interaction with polymerized collagen, decreased nuclear HDAC4 levels suppress miR-29c transcription, thereby activating type I collagen expression.

Deidentified patient samples were obtained under a waiver of informed consent from the University of Minnesota Institutional Review Board.

Eleven primary mesenchymal cell lines were established from patients with IPF. All patients fulfilled the criteria for the diagnosis of IPF as established by the American Thoracic Society and the European Respiratory Society (8). Diagnosis of IPF was confirmed by microscopic analysis of lung tissue, which demonstrated the characteristic morphological findings of usual interstitial pneumonia. Cell lines were derived from lungs removed at the time of transplantation or death as described (31). Patient controls were selected to be similar in age to patients with IPF with nonfibrotic lung disorders. Ten nonfibrotic primary control adult human lung fibroblast lines were used. These lines were established from lung tissue uninvolved by the primary disease process: adenocarcinoma (n = 4), squamous cell carcinoma (n = 1), carcinoid tumor (n = 2), fibrosarcoma (n = 1), leimyosarcoma (n = 1), or bronchiectasis (n = 1). Primary lung mesenchymal cell lines were generated by explant culture and/or mechanical/chemical dispersion and maintained in high-glucose DMEM containing 10% FCS. Cells were used from passages 3, 4, and 5.

Phosphorylated and HDAC4 antibodies were obtained from Cell Signaling (Danvers, MA). Collagen I antibodies were obtained from AbD Serotec (Raleigh, NC) and Southern Biotech (Birmingham, AL). PP2Ac monoclonal antibody was obtained from Millipore (Temecula, CA).

Western blot analysis was performed on cell lysates as described (16–18).

Quantitative RT-PCR was performed as described (32). The primer sequences were as follows: COL1A2 forward: 5′TGCCTAGCAACATGCCAATC COL1A2 reverse: 5′TGAGCAGCAAAGTTCCCACC HDAC4 forward: 5′TCCAGATGGACTTTCTGGCCG HDAC4 reverse: 5′GCTGGGCATGTGGTTCACG miR-29c: 5′UAGCACCAUUUGAAAUCGGUUA.

Gain of miR-29c function was performed as described (32).

Adenoviral vectors containing wild-type HDAC4 (Ad-HDAC4), wild-type PP2Ac (Ad-PP2Ac), and control (Ad-GFP) constructs were amplified to high titer according to the manufacturer’s instructions (abm, Richmond, BC, Canada). Cells were infected with adenoviral vectors at a multiplicity of infection of 1:20.

Lentivirus plasmids (HDAC4 V2LHS_239051 Mature antisense: ACAATGAAGAAATGGTTTC) were obtained from Thermo Scientific (Waltham, MA). Viral supernatants were harvested 48 hours after transfection, filtered through a 0.45-μm pore size polyvinylidene fluoride filter, aliquoted, and frozen at −80°C. IPF fibroblasts were plated in six-well dishes and incubated overnight. Virus was added to cells in DMEM–10% FBS with polybrene (final concentration, 8 μg/ml). The plate was centrifuged (1,200 × g 25°C) for 1 hour and then incubated for 16 hours (37°C, 5% CO2). The virus was removed, 2 ml fresh medium per well was added, and incubation continued for 48 hours.

Immunohistochemistry was performed on 4-μm–sectioned, paraffin-embedded lung tissue specimens as described using a monoclonal antibody to HDAC4 (1:100) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and a biotinylated horse anti-mouse secondary antibody (1:500) (18).

Comparisons of data among experiments were performed with the unipolar unpaired or paired Student’s t test. Experiments were independently replicated a minimum of three times. Data are expressed as mean ± SD. P < 0.05 was considered significant.

A key pathological feature of the IPF fibroblast phenotype is increased synthesis and deposition of type I collagen, leading to expansion of the alveolar wall and obliteration of the alveolar airspace (16–18, 33). Previous work indicates that miR-29, which regulates the expression of a variety of extracellular matrix components including type I collagen, is suppressed in IPF (15). This implicates a vital role for miR-29 in the excessive collagen production characteristic of IPF. However, the mechanism for suppression of miR-29 in IPF remains unclear. Because IPF fibroblasts are the key effector cell of the fibrotic response, synthesizing and depositing type I collagen, we began our experiments by quantifying miR-29 and type I collagen levels when IPF fibroblasts interact with polymerized collagen, which is a major component of the IPF fibrotic matrix. When IPF fibroblasts were seeded on polymerized collagen matrices, they expressed increased levels of collagen I and lower levels of miR-29 compared with control fibroblasts ( Figure 1A ). We next examined whether the increased expression of type I collagen in IPF fibroblasts was influenced by their interaction with the polymerized collagen matrix. To address this issue, we seeded IPF fibroblasts on polymerized collagen matrices or tissue culture plastic and examined collagen I and miR-29 expression. IPF fibroblasts seeded on polymerized collagen matrices expressed higher levels of collagen I and lower levels of miR-29 compared with when they were seeded on tissue culture plastic ( Figure 1B ). This suggested that IPF fibroblast interaction with the polymerized type I collagen matrix results in increased collagen I expression due to aberrant regulation of miR-29 expression.

Figure 1. Type I collagen expression is high and microRNA (miR)-29c is low in idiopathic pulmonary fibrosis (IPF) fibroblasts. (A) Primary IPF (n = 6) and control human lung fibroblasts (n = 4) were seeded on polymerized type I collagen matrices for 4 hours. Left panel: Type I collagen protein expression was quantified by Western blot analysis. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) is shown as a loading control. Right panel: miR-29c levels were quantified by quantitative RT-PCR (qRT-PCR) and are shown as the ratio of miR-29 to U6-2 small nuclear RNA (RNU) expression. (B) IPF fibroblasts (n = 3) were plated on tissue culture plastic (TC) or polymerized type I collagen matrices (PC). Left panel: Type I collagen expression was quantified by Western analysis. GAPDH is shown as a loading control. Right panel: miR-29c levels were quantified by qRT-PCR and are shown as the ratio of miR-29 to RNU expression.

We have previously found that the α2β1 integrin/PP2A axis is abnormal during IPF fibroblast interaction with polymerized type I collagen and that this axis is a key regulator of the IPF fibroblast phenotype (18). Low α2β1 integrin expression results in a failure of PP2A to be appropriately activated when IPF fibroblasts interact with polymerized collagen (18). PP2A is a critical regulator of HDAC4 function (21). HDAC4 contains an N-terminal regulatory domain that is subject to phosphorylation. When phosphorylated, HDAC4 is susceptible to degradation by the proteasome. However, when dephosphorylated by PP2A, HDAC4 protein is stable (21–24). Importantly, HDAC4 has been linked to regulation of miR-29 expression (30). Because PP2A function is abnormal in IPF, we hypothesized that aberrant PP2A regulation of HDAC4 function may lead to suppression of miR-29. To begin to analyze HDAC4 function in IPF fibroblasts, we first examined HDAC4 expression in IPF fibroblasts seeded on polymerized type I collagen matrices as a function of time. Four hours after seeding cells on polymerized collagen, IPF fibroblast HDAC4 protein levels were decreased. By 24 hours, HDAC4 protein levels were markedly decreased (86% decrease). In contrast, although HDAC4 expression also decreased in control fibroblasts as a function of time, the level of decrease (47% decrease) was more modest ( Figure 2A ). We also quantified the level of phosphorylated HDAC4 in IPF and control fibroblasts seeded on polymerized collagen matrices as a function of time. We found that at the 4-hour time-point the level of phosphorylated HDAC4 had increased approximately 1.7-fold in IPF fibroblasts compared with a 47% increase in control fibroblasts. Phosphorylated HDAC4 is prone to degradation by the proteasome. Therefore, we next examined the effect of pretreating IPF fibroblasts with the proteasome inhibitor MG-132 on HDAC4 protein levels when seeded on polymerized collagen. We found that HDAC4 protein levels were relatively preserved in IPF fibroblasts treated with MG-132 compared with vehicle control ( Figure 2B ). Together, these data indicate that HDAC4 becomes phosphorylated and degraded when IPF fibroblasts interact with polymerized collagen.

Figure 2. Histone deacetylase (HDAC) 4 protein levels are decreased in IPF fibroblasts. (A) IPF and control lung fibroblasts were seeded on polymerized type I collagen matrices as a function of time. Top panel: Phosphorylated HDAC4 (pHDAC4) and HDAC4 expression were examined by Western blot analysis. GAPDH is shown as a loading control. Bottom panel: pHDAC4 and HDAC4 levels were quantified by densitometry. (B) IPF fibroblasts were pretreated with the proteasome inhibitor MG132 (20 nM 60 min) or vehicle control (ethanol [ETOH]) and seeded on polymerized collagen for 0, 2, 4, or 8 hours. HDAC4 protein levels were quantified by Western blot analysis. GAPDH is shown as a loading control. All experiments were repeated with three independent cell lines to confirm the findings.

We have found that phospho-HDAC4 levels are high in response to IPF fibroblast interaction with polymerized collagen, whereas PP2A function is low (18). Because PP2A binds and dephosphorylates HDAC4, thereby promoting HDAC4 nuclear import (21), we next sought to determine whether we could detect a physical association of PP2A with HDAC4 when control or IPF fibroblasts were seeded on polymerized collagen matrices for up to 4 hours. Although a PP2A/HDAC4 complex could be readily detected in control fibroblasts, this complex was below the limits of detection in IPF fibroblasts ( Figure 3A ). Furthermore, when we pretreated control fibroblasts with the PP2A inhibitor okadaic acid, the level of phosphorylated HDAC4 increased sharply ( Figure 3B ). Taken together, these data suggest that, when IPF fibroblasts interact with polymerized type I collagen, low PP2A phosphatase function leads to increased levels of phosphorylated HDAC4.

Figure 3. Low protein phosphatase (PP) 2A in IPF fibroblasts results in HDAC4 hyperphosphorylation and decreases its nuclear localization. (A) IPF and control fibroblasts were plated on type I polymerized collagen for 4 hours. Immunoprecipitation of HDAC4 was performed, and samples were analyzed for association with PP2Ac. Immunoprecipitation with isotype antibody was used as a control. (B) Control lung fibroblasts were pretreated with the PP2A inhibitor okadaic acid (OA) (10 nM 60 min) or DMSO as a control. The cells were then seeded on type I polymerized collagen matrices and phosphorylated, and HDAC4 protein expression was examined by Western blot analysis as a function of time. GAPDH is shown as a loading control. (C) IPF and control fibroblasts were seeded on polymerized collagen for 4 hours. The cells were lysed, and nuclear and cytoplasmic fractions were analyzed for HDAC4 expression by Western blot analysis. Lamin A/C is shown as a nuclear loading control GAPDH is shown as a cytoplasmic loading control. (D) IPF and control fibroblasts were seeded on polymerized type I collagen for 4 hours. The cells were stained with HDAC4 antibody conjugated with Cy-3. 4′6-Diamidino-2-phenylindole (DAPI) indicates nuclear staining. Arrows point to four IPF fibroblasts with low nuclear HDAC4 expression. (E) PP2Ac was overexpressed in IPF fibroblasts using an adenoviral vector containing a wild-type PP2Ac construct (Ad-PP2Ac). Cells infected with empty vector served as control (Ad-EV). The cells were seeded on polymerized collagen for 4 hours and then lysed. Nuclear (N) and cytoplasmic (C) fractions were analyzed for PP2Ac and HDAC4 expression by Western blot analysis. Lamin A/C is shown as a nuclear loading control GAPDH is shown as a cytoplasmic loading control.

HDAC4 regulates gene expression by undergoing nucleocytoplasmic shuttling in response to external cues. Nucleocytoplasmic shuttling of HDAC4 is controlled by its phosphorylation state (21). Dephosphorylation by PP2A stabilizes HDAC4 and promotes its nuclear accumulation, whereas phosphorylated HDAC4 resides in the cytoplasm, where it is subject to proteasomal degradation (21, 22). Therefore, we next examined HDAC4 localization in IPF and control fibroblasts seeded on polymerized collagen matrices. We found that nuclear HDAC4 levels were markedly decreased in IPF fibroblasts compared with control ( Figure 3C ). In addition, cytoplasmic HDAC4 levels were decreased. After 4 hours on polymerized collagen, nuclear HDAC4 levels in IPF fibroblasts were at the lower limits of detection. Consistent with this, immunocytochemistry demonstrated that HDAC4 was predominantly cytoplasmic in IPF fibroblasts seeded on polymerized collagen matrices ( Figure 3D ). In contrast, HDAC4 was found in the nucleus and cytoplasm in control fibroblasts ( Figures 3C and 3D ). To directly test the role of PP2Ac in regulating HDAC4 localization in IPF fibroblasts, we used an adenoviral vector to overexpress PP2A in IPF fibroblasts. In accord with a direct causal role, PP2Ac gain-of-function in IPF fibroblasts augmented nuclear HDAC4 levels ( Figure 3E ). Thus, our data indicate that, when IPF fibroblasts interact with polymerized collagen, aberrantly low PP2A function results in HDAC4 phosphorylation and decreased levels of HDAC4 in the nucleus and cytoplasm.

We next examined the role of HDAC4 in regulating type I collagen expression in our system. To do this, we performed HDAC4 gain-of-function experiments in IPF fibroblasts. Overexpression of HDAC4 increased nuclear and cytoplasmic HDAC4 protein levels ( Figure 4A ). Importantly, we found that overexpression of HDAC4 in IPF fibroblasts markedly decreased type I collagen mRNA and protein expression ( Figure 4B ) while increasing miR-29 expression ( Figure 4C ). In contrast, knockdown of HDAC4 by short hairpin RNA (shRNA) in control fibroblasts decreased HDAC4 expression by 40% and substantially increased type I collagen mRNA and protein expression ( Figure 4D ) while decreasing miR-29 expression ( Figure 4E ). These data suggest that HDAC4 regulates collagen 1 expression via miR-29.

Figure 4. HDAC4 regulates type I collagen and miR-29 expression. (AC) HDAC4 was overexpressed in IPF fibroblasts using an adenoviral vector containing a wild-type HDAC4 construct (Ad-HDAC4). Controls consisted of cells expressing empty vector (Ad-EV). (A) Cells were seeded on polymerized type I collagen matrices for 4 hours. Cells were lysed and nuclear (N) and cytoplasmic (C) fractions were analyzed for HDAC4 protein levels by Western analysis (left panel). Nuclear and cytoplasmic HDAC4 levels were quantified by densitometric analysis (middle and right panels). (B) COL1A2 messenger RNA (mRNA) (left panel) and collagen I protein (right panel) expression were examined by qRT-PCR and Western blot analysis, respectively. GAPDH is shown as a loading control. (C) miR-29c levels were quantified by qRT-PCR. (D and E) HDAC4 was knocked down in control fibroblasts using a lentiviral vector containing HDAC4 short hairpin RNA (shRNA). Cells infected with lentiviral vector containing scrambled shRNA (Scr-shRNA) were used as a control. The cells were plated on polymerized type I collagen for 4 hours. (D) COL1A2 mRNA (left panel) and collagen I and HDAC4 protein levels (top right panel) were examined by qRT-PCR and Western blot analysis, respectively. GAPDH is shown as a loading control. Collagen I and HDAC4 protein levels were quantified by densitometric analysis (bottom right panel). (E) miR-29c levels were quantified by qRT-PCR.

We have shown that PP2A gain-of-function increases HDAC4 levels in IPF fibroblasts ( Figure 3E ) and that high levels of HDAC4 decrease collagen I expression ( Figure 4B ). We next sought to analyze the role of the PP2A/HDAC4 axis in regulating collagen I and miR-29 expression. We found that PP2A gain-of-function in IPF fibroblasts decreased collagen I mRNA expression while increasing miR-29 expression ( Figure 5A , lane 3 and Figure 5B, lane 3). To ascertain whether PP2A-mediated increase in HDAC4 is responsible for the changes in collagen I and miR-29 expression, we knocked down HDAC4 in IPF fibroblasts in which PP2A had been overexpressed. We found that knockdown of HDAC4 in IPF fibroblasts overexpressing PP2A resulted in increased collagen I expression and suppression of miR-29 expression ( Figure 5A , lane 4 and Figure 5B , lane 4). These data indicate that PP2A regulates collagen I and miR-29 expression via HDAC4 and strongly suggest that in IPF fibroblasts, low PP2A function leads to decreased HDAC4 levels. Low HDAC4 function in turn suppresses miR-29, which leads to an increase in collagen I expression.

Figure 5. PP2A/HDAC4 axis regulates type I collagen expression via miR-29. (A and B) HDAC4 was knocked down by HDAC4 shRNA in IPF fibroblasts in which PP2Ac was overexpressed using an adenoviral vector (Ad-PP2Ac/HDAC4-shRNA). Controls consisted of IPF fibroblasts in which PP2A was overexpressed and treated with scrambled shRNA (Ad-PP2Ac/Scr-shRNA), cells infected with empty vector (control for PP2A) and treated with HDAC4-shRNA (Ad-EV/HDAC4-shRNA), and cells infected with empty vector and treated with scrambled shRNA (Ad-EV/Scr-shRNA). qRT-PCR was done to assess mRNA levels of Col1A1 (A) and miR-29c (B). (C and D) Control fibroblasts in which HDAC4 had been knocked down were infected with a lentiviral vector containing a miR-29c construct to overexpress miR-29 (HDAC4-shRNA/miR-29 overexpression [OE]). Controls consisted of control fibroblasts in which HDAC4 had been knocked down and then treated with empty vector (HDAC4-shRNA/EV), control fibroblasts treated with scrambled shRNA in which miR-29c had been overexpressed (Scr-shRNA/miR-29 OE), and control fibroblasts treated with scrambled shRNA and empty vector (Scr-shRNA/EV). The cells were seeded on polymerized type I collagen matrices for 4 hours. (C) Left panel: To confirm HDAC4 knockdown, HDAC4 expression was quantified by Western blot analysis. Shown is the ratio of HDAC4 to GAPDH. Right panel: To confirm miR-29 overexpression, miR-29c expression was quantified by qRT-PCR. Shown is the ratio of miR-29 to RNU. EV, empty vector. (D) Type I collagen expression was assessed by Western blot analysis. GAPDH is shown as a loading control (top panel). Collagen I protein expression was quantified by densitometric analysis (bottom panel).

To directly test whether the regulation of collagen I by HDAC4 operates through its effect on miR-29c, we conducted a rescue experiment where we overexpressed miR-29c in control fibroblasts in which HDAC4 had been knocked down. Knockdown of HDAC4 by shRNA decreased HDAC4 expression by 45% ( Figure 5C , left panel). Overexpression of miR-29c increased miR-29 levels by 65% ( Figure 5C , right panel). We found that knockdown of HDAC4 increased collagen I expression ( Figure 5D , lane 2), whereas overexpression of miR-29c suppressed collagen I expression ( Figure 5D , lane 3). When the miR-29c precursor was ectopically expressed in control fibroblasts in which HDAC4 had been knocked down, the previously observed increase in type I collagen was ablated ( Figure 5D , lane 4). This places miR-29 suppression of type I collagen downstream of HDAC4. Taken together, these data support a model in which the failure of IPF fibroblasts to appropriately activate PP2A when interacting with the polymerized collagen matrix leads to decreased nuclear HDAC4 levels. Depleted nuclear HDAC4 in turn suppresses miR-29 expression and subsequently activates type I collagen expression.

To examine the in vivo relevance of our findings, we analyzed HDAC4 expression in IPF lung tissue specimens by immunohistochemistry. We found a paucity of HDAC4 immunoreactive cells within IPF fibroblastic foci ( Figure 6A ). In contrast, numerous HDAC4 immunoreactive cells were interspersed within the walls of anatomically normal alveolar structures in control lung tissue. Importantly, many cells demonstrated nuclear HDAC4 staining ( Figure 6C ). As controls for HDAC4 immunohistochemistry, no immunoreactivity was present in IPF or control lung tissue stained with secondary antibody only ( Figures 6B and 6D , respectively). Thus, HDAC4 levels are low in IPF fibroblasts populating the IPF fibrotic reticulum.

Figure 6. IPF fibroblastic foci contain a paucity of HDAC4 immunoreactive cells. IPF (A and B) and control (C and D) human lung tissue specimens (n = 3 each) were analyzed for HDAC4 immunoreactivity by immunohistochemistry. (A) Representative image of immunohistochemistry of IPF lung tissue demonstrating a paucity of HDAC4-expressing cells in the IPF fibrotic reticulum. (B) Negative control: IPF lung tissue stained with secondary antibody only. (C) Numerous HDAC4-expressing cells were detected in alveolar structures of anatomically normal control lung tissue. Arrows point to HDAC4 immunoreactive cells. Inset: High-power image of a cell displaying nuclear HDAC4 immunoreactivity. (D) Negative control: control lung tissue stained with secondary antibody only. Scale bar, 50 μm.

Fibroblasts derived from the lungs of patients with IPF display a distinct pathologic phenotype, including increased expression of type I collagen. Prior studies have demonstrated that miR-29, a key regulator of collagen expression, is abnormally suppressed in IPF (reviewed in Reference 15). However, the mechanism for this aberrant suppression of miR-29 in IPF remained unclear. In this report, we demonstrate that, in response to IPF fibroblast interaction with polymerized type I collagen, an aberrant PP2A/HDAC4 axis leads to hyperphosphorylation of HDAC4 and reduced HDAC4 nuclear import, resulting in suppression of miR-29 and activation of type I collagen expression. These data suggest an important role for the PP2A/HDAC4/miR-29 axis in regulating the excessive collagen production that is a hallmark feature of IPF.

We have recently discovered that, in response to fibroblast residence on decellularized IPF fibrotic matrices, miR-29 expression is suppressed, causing increased type I collagen expression (32). These findings indicate that fibroblasts manifest a pathologic feed-forward circuit in which residence on the fibrotic matrix promotes the production of excessive amounts of type I collagen, driving progressive fibrosis. However, the mechanism by which the fibrotic matrix confers pathologic suppression of miR-29 and abnormally high expression of type I collagen remained unclear. Here we show that culture of IPF fibroblasts on polymerized type I collagen matrices mirrored the response of culture on decellularized IPF matrices in terms of suppression of miR-29 levels and augmentation of collagen expression. Thus, aberrant control of miR-29–type I collagen feedback by the IPF fibroblast was recapitulated on polymerized type I collagen matrices. This suggests that IPF fibroblast interaction with type I collagen in the IPF fibrotic matrix is a major determinant regulating the pathologic suppression of miR-29 in IPF.

The IPF extracellular matrix is rich in cross-linked type I collagen, which is mechanically altered, displaying increased stiffness. Recent studies indicate that the fibrotic collagen-rich matrix can activate fibroblasts and drive fibrosis (32, 34–38). Importantly, studies in cancer indicate that a cross-linked type I collagen matrix alters integrin signaling and promotes tumor progression (39–41). α2β1 integrin is a major collagen receptor that cells utilize when adhering to collagen. Interestingly, many cancer daughter cells display abnormally low levels of α2β1 integrin, and this has been linked to cancer progression and metastasis, suggesting that aberrant α2β1 integrin/type I collagen interaction promotes acquisition of a pathologic phenotype (42–46). In further support of this concept, a recent study demonstrated that fibroblasts cultured for a prolonged period on matrices with altered mechanical properties displayed an activated phenotype that persisted even when the cells were returned to pliable matrices (47). This suggests that a fibrotic, noncompliant matrix is capable of “reprogramming” fibroblasts such that they manifest a durable pathologic phenotype. The mechanism by which a fibrotic type I collagen matrix “reprograms” fibroblasts remains to be elucidated.

Recent studies have begun to uncover the basic mechanism(s) by which α2β1/collagen interaction regulates cellular phenotype. Seminal work by Heino and colleagues discovered that when α2β1 integrin binds polymerized type I collagen, the PP2A phosphatase becomes activated (19). We have previously found that α2β1 integrin expression is low during IPF fibroblast interaction with polymerized type I collagen matrices and that this results in a failure for PP2A to be appropriately activated (18). Importantly, prior work demonstrated that PP2A is a critical regulator of HDAC4, a class II histone deactylase that has been shown to regulate normal fibroblast differentiation (19). HDAC4 function is controlled by its phosphorylation state (21). HDAC4 contains an N-terminal regulatory domain that is subject to phosphorylation. When phosphorylated, HDAC4 is retained in the cytoplasm, where it is prone to degradation by the proteasome. However, when dephosphorylated by PP2A, HDAC4 undergoes nuclear import. Within the nucleus, deacetylation of histones by HDACs results in chromatin compaction, which represses gene transcription, altering cell phenotype. Importantly, studies indicate that the matrix microenvironment can reprogram cells through epigenetic regulatory mechanisms, including histone modifications via HDACs, further supporting the concept that the collagen-rich IPF fibrotic matrix may regulate the IPF fibroblast phenotype via HDAC4.

Here we demonstrate that when IPF fibroblasts interact with polymerized collagen, low PP2A activity causes HDAC4 to be phosphorylated and degraded and HDAC4 protein levels are markedly depressed. This results in depleted nuclear HDAC4. Moreover, we have found that in IPF fibroblasts restoration of PP2A function increases nuclear HDAC4 levels and miR-29 expression while decreasing collagen I levels. Our studies demonstrate that HDAC4 functions downstream of PP2A in regulating miR-29 and collagen I expression. We have discovered that HDAC4 regulates miR-29 expression. We show that knockdown of HDAC4 suppresses miR-29 levels and promotes type I collagen expression, whereas overexpression of HDAC4 increases miR-29 expression while suppressing collagen expression. miR-29 is a potent post-translational inhibitor of type I collagen. Interestingly, a recent study involving mouse hepatic stellate cells also found a relationship between HDAC4 and miR-29 (30). However, in contrast to our findings, this study found that inhibition of HDAC4 resulted in up-regulation of miR-29. The reason for these discrepant results is not clear, but one possible explanation is that HDAC4 regulation of miR-29 is cell type and condition specific.

Our data support a model where IPF fibroblast interaction with the polymerized collagen matrix results in a failure of PP2A to be activated, which depletes nuclear HDAC4 levels. Decreased nuclear HDAC4 suppresses miR-29 expression, which in turn causes an increase in COL1A1 and COL1A2 translation and subsequently collagen I protein expression. To verify this, we performed a series of experiments. First, we demonstrated that overexpression of HDAC4 in IPF fibroblasts augments miR-29 expression while decreasing type I collagen protein levels. Next, we showed that knockdown of HDAC4 in control fibroblasts suppressed miR-29 levels and increased collagen expression. We went on to demonstrate that the increased collagen expression that resulted by knocking down HDAC4 can be reversed by gain-of-function of miR-29. These data strongly support the concept that, in response to interaction with polymerized type I collagen, the PP2AHDAC4 axis regulates the increased production of type I collagen, a hallmark feature of the IPF fibroblast phenotype, by suppressing miR-29 expression.

Precisely how HDAC4 regulates miR-29 expression is unclear. Prior work indicates that upon nuclear import, HDAC4 functions by binding to tissue-specific transcription factors and inhibiting their function (22, 25–29). Transcription factors can be positive or negative regulators of gene expression. If HDAC4 binds to a transcription factor that is a positive gene regulator, it will lead to inhibition of gene expression. In contrast, if HDAC4 binds to a transcription factor that is a negative regulator of gene expression, it will lead to activation of gene transcription. Thus, one possible mechanism by which HDAC4 regulates miR-29 involves HDAC4 binding and inhibiting a transcription factor that functions as a negative regulator of miR-29, possibly Sp1, which has been shown to bind to a regulatory element on the miR-29 gene and to repress its transcription (48). In this scenario, the repressive function of the transcription factor toward miR-29 is released, thereby increasing miR-29 gene transcription. We acknowledge a key limitation of our study because we were unable to identify the transcription factor responsible for regulating miR-29 expression. Our future studies will focus on identifying this elusive transcription factor, which will further strengthen our understanding of type I collagen regulation in IPF fibroblasts.

In summary, we demonstrate that during IPF fibroblast interaction with polymerized collagen, inappropriately low PP2A activity leads to aberrant HDAC4 function. This altered HDAC4 function is responsible for the pathologic decrease in miR-29c and acquisition of a hallmark feature of the IPF fibroblast: increased type I collagen protein translation and expression. Our data suggest that IPF fibroblasts have been epigenetically reprogrammed to respond pathologically when they interact with polymerized type I collagen matrices and that the α2β1 integrin/PP2A/HDAC4/miR-29c axis underlies this pathologic response.


This work was supported by grants from the National Institutes of Health/National Cancer Institute (NIH/NCI) U01 CA154969 (SKP, MD), U54 CA209971 (SKP). AJG received support from U24 CA224309 (NCI) and the Fund for Cancer Informatics. The Genome Sequencing Service Center provided by Stanford Center for Genomics and Personalized Medicine Sequencing Center is supported by NIH grant S10OD02014. The Redcap clinical databased used in this project is supported by grant UL1 TR001085 from NIH/NCRR.

Affiliations

Department of Medicine (Biomedical Informatics Research), Stanford University, Stanford, CA, 94035, USA

Andrew J. Gentles, Armon Azizi & Alice Yu

Department of Biomedical Data Science, Stanford University, Stanford, CA, 94035, USA

Andrew J. Gentles & Sylvia K. Plevritis

Cancer Institute, Stanford University, Stanford, CA, 94035, USA

Andrew J. Gentles, Angela Bik-Yu Hui, Armon Azizi, Youngtae Jeong, Maximilian Diehn & Sylvia K. Plevritis

Department of Radiation Oncology, Stanford University, Stanford, USA

Angela Bik-Yu Hui, Youngtae Jeong & Maximilian Diehn

Stem Cell Institute, Stanford University, Stanford, USA

Angela Bik-Yu Hui, Youngtae Jeong & Maximilian Diehn

Translational Medicine Group, Abbvie, North Chicago, IL, 60064, USA

Stanford Center for Genomics and Personalized Medicine, Stanford, USA

Department of Radiology, Stanford University, Stanford, USA

Gina Bouchard, David A. Knowles, Alborz Bejnood & Sylvia K. Plevritis

Department of Computer Science, Columbia University, New York, USA

Department of New Biology, Daegu Gyeongbuk Institute of Science and Technology (DGIST), Daegu, South Korea

Broad Institute of MIT and Harvard, Cambridge, USA

Department of Pathology, Stanford University, Stanford, USA

Erna Forgó, Sushama Varma, Rob West & Matt van de Rijn

Department of Thoracic Surgery, Stanford University, Stanford, USA

Department of Gynecology Oncology, CAP/Stanford Health Care Tri-Valley Outreach, Pleasanton, CA, USA

Division of Pulmonary & Critical Care, Stanford University, Stanford, USA

Moffitt Cancer Center and the University of South Florida Division of Pulmonary & Critical Care Medicine/Department of Oncologic Sciences, Tampa, FL, USA

Thoracic and Oncologic Surgery Branch, CCR, NIH – National Cancer Institute, Bethesda, MD, USA

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

Contributions

AJG, MD, and SKP conceived and performed analysis and supervised the project. CDH obtained primary lung tumors from the operating room during surgery. AH, WF, GB, YJ, EF, SV, and YX performed wet lab experiments and validation. AA, RVN, DAK, AY, AB, and AJG conducted the computational analyses and LTMI website construction. AK and VSN extracted and curated clinical information from patient records. RW and MvdR supervised and oversaw immunohistochemistry experiments and validation. AJG, MD, and SKP wrote the paper with contributions from all authors. All authors approved submission of the final manuscript.

Corresponding authors


Materials and methods

Infection of lungs

Mice were infected at 4–6 weeks of age. AdCre:CaPi coprecipitates were prepared as described (Fasbender et al. 1998). Mice were anesthetized with avertin. AdCre:CaPi coprecipitates were administered intranasally in two 62.5-μL instillations. The second instillation was administered when breathing rates had returned to normal.

Molecular and biochemical analysis

For verification of Cre-mediated recombination, DNA was prepared from portions of lungs, tails, and kidneys. PCR was performed with primers flanking the Lox–Stop–Lox cassette (sequence available upon request). The K-ras and Lox-K-ras G12D alleles were detected, yielding a 265-bp and a 305-bp product, respectively.

For protein analysis of lungs by IP-Western, tissue lysates and immunoprecipitation were performed as described in Johnson et al. (1997). Western analysis was performed as described in Johnson et al. (2001).

Histological analysis and immunohistochemistry

Animals were killed at the times indicated and subjected to full necropsy. Histological and immunohistochemical analyses were performed as described in Johnson et al. (1997).

For immunofluorescence, sections were blocked at room temperature for 2 h in PBS containing 0.2% Triton X-100 and normal horse serum. They were then incubated at 4°C overnight with both rabbit polyclonal anti-CCA at 1:1000 and goat polyclonal anti-pro-SP-C (cat #RDI-RTSURFCCabG) at 1:100 diluted in blocking buffer. Following incubation, sections were washed 3 times for 5 min with 0.2% Triton X-100 in PBS. Sections were then incubated at room temperature with rhodamine-conjugated anti-goat at a 1:200 dilution for 30 min followed by a 30-min incubation with FITC-conjugated anti-rabbit used at a 1:1000 dilution. Sections were then counterstained for 5 min with DAPI.


Materials and methods

A list of the sources of all the datasets used in this study can be found in S1 Table.

Building the datasets of SARS-CoV-2 target host genes/proteins

We assembled 4 SARS-CoV-2 datasets of target host genes/proteins: (1) 246 DEGs in human bronchial epithelial cells infected with SARS-CoV-2 [21] (GSE147507), denoted as SARS2-DEG (2) 293 DEPs in human Caco-2 cells infected with SARS-CoV-2 [22], denoted as SARS2-DEP (3) 134 strong literature-evidence-based pan-human coronavirus target host proteins from our recent study [30] with 15 newly curated proteins, denoted as HCoV-PPI and (4) 332 proteins involved in PPIs with 26 SARS-CoV-2 viral proteins identified by affinity purification–mass spectrometry (AP-MS) [8], denoted as SARS2-PPI. Finally, due to the interactome nature of HCoV-PPI and SARS2-PPI, we combined these datasets as the fifth SARS-CoV-2 dataset, which has 460 proteins and is denoted as PanCoV-PPI. Details of these datasets can be found in S2 Table.

SARS2-DEG.

In the original study, the primary human bronchial epithelial cells were infected with SARS-CoV-2 for 24 hours. The transcriptome profiles of infected (3 replicates) and uninfected cells (3 replicates) were characterized, and the fold change (FC) and FDR for each gene were calculated by DESeq2 and provided in the original study. We applied a cutoff of |log2FC| > 0.5 and FDR < 0.05 to identify the DEGs.

SARS2-DEP.

As described in the previous study [22], human Caco-2 cells were infected with SARS-CoV-2 for up to 24 hours. Proteomics assays of the infected and uninfected cells were measured at 24 hours in triplicates. We used the results at 24 hours, as the original study showed most DEPs at 24 hours. The P values were computed using 2-sided unpaired Student t tests with equal variance assumed in this study. We converted the P value to FDR using the “fdrcorrection” function in the Python package statsmodels v0.11.1 and used a cutoff of FDR < 0.05 to identify the DEPs.

Collection of 4 additional virus–host gene/protein networks

To characterize the SARS-CoV-2 datasets, we downloaded 4 virus–host gene/protein networks from previous studies for comparison: (1) 900 virus–host interactions identified by gene-trap insertional mutagenesis connecting 10 other viruses and 712 host genes [35] (2) 2,855 virus–host interactions identified from RNAi connecting 2,443 host genes and 55 pathogens [35] (3) 579 host proteins mediating translation of 70 innate immune-modulating viORFs [36] and (4) 1,292 host genes identified by Co-IP+LC/MS that mediate influenza–host interactions [37]. All details for these 4 virus–host gene/protein networks are provided in S3 Table.

Building the disease gene profiles

We compiled the disease-associated gene sets from various sources. All databases were accessed on March 26, 2020.

Cancer.

We defined a driver gene as a gene that had significantly enriched driver mutations based on whole-genome or whole-exome sequencing data or reported experimental data from the Cancer Gene Census [42,43] or the original publications from The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/). The pan-cancer driver genes were retrieved from the Cancer Gene Census [42,43]. Driver genes for individual cancer types were from a previous study [44].

Mendelian disease genes (MDGs).

A set of 2,272 MDGs were retrieved from the Online Mendelian Inheritance in Man (OMIM) database [106].

Orphan disease-causing mutant genes (ODMGs).

A set of 2,124 ODMGs were retrieved from a previous study [107].

Cell cycle genes.

A set of 910 human cell cycle genes were downloaded from a previous study in which they were identified by a genome-wide RNAi screening [108].

Innate immune genes.

A set of 1,031 human innate immune genes were collected from InnateDB [109].

Genes associated with autoimmune, pulmonary, neurological, cardiovascular, and metabolic diseases.

The disease-associated genes/proteins were extracted from HGMD [45]. HGMD is a well-documented database, and we downloaded the whole database for data analysis and extraction using well-documented disease ontology terms [46]. We defined a disease-associated gene as a gene that has at least 1 disease-associated mutation in original publications provided in HGMD. The details, including the sources, number of genes, mutations associated with the disease, and terms used to identify diseases in HGMD, are provided in S4 Table.

Functional enrichment analysis

We performed Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) biological process enrichment analyses to reveal the biological relevance and functional pathways of the 5 SARS-CoV-2 datasets. All functional enrichment analyses were performed using Enrichr [110]. An overview of the virus infection-related pathways and ontology terms shared by 1 or more datasets was generated by searching for significant pathways or terms (FDR < 0.05). The enrichment analysis results for the 5 SARS-CoV-2 gene/protein sets can be found in S1–S5 Figs.

Selective pressure and evolutionary rate characterization

We calculated the dN/dS ratio [111] and the evolutionary rate ratio [112] as described in our previous study [113]. A dN/dS ratio below, equal to, or above 1 suggests purifying selection, neutral evolution, or positive Darwinian selection, respectively [114]. The evolutionary rate ratio was computed using the criterion that a ratio > 1 indicates a fast rate and a ratio < 1 indicates a slow rate [112]. The dN/dS and evolutionary rate ratios of the genes in the 5 SARS-CoV-2 datasets and 4 additional virus gene/protein sets can be found in S2 and S3 Tables.

Tissue specificity analysis

The RNA-Seq data (transcripts per million [TPM]) of 33 tissues from the GTEx V8 release (accessed on March 31, 2020 https://www.gtexportal.org/home/) were downloaded. Genes with count per million (CPM) ≥ 0.5 in over 90% of samples in a tissue were considered tissue-expressed genes, and otherwise tissue-unexpressed. To quantify the expression specificity of gene i in tissue t, we calculated the mean expression Ei and the standard deviation σi of a gene’s expression across all considered tissues. The significance of gene expression specificity in a tissue is defined as (1)

Risk ratio analysis for COVID-19 patients

PubMed, Embase, and medRxiv databases were searched for publications as of April 25, 2020 (S18 Fig). The search was limited to articles in English describing the demographic and clinical features of SARS-CoV-2 cases. We used the search term (“SARS-COV-2” OR “COVID-19” OR “nCoV 19” OR “2019 novel coronavirus” OR “coronavirus disease 2019”) AND (“clinical characteristics” OR “clinical outcome” OR “comorbidities”). Only research articles were included reviews, case reports, comments, editorials, and expert opinions were excluded. Three criteria were used to select studies from a total of 1,054 initial hits: (1) studies that had ≥20 COVID-19 patients (2) studies that grouped the outcomes by degree of severity of COVID-19 (e.g., severe versus non-severe) according to the American Thoracic Society guidelines for community-acquired pneumonia and (3) studies that were from different institutions. Two criteria were used for exclusion: (1) studies that focused on specific populations (e.g., only death cases, pregnant women, children, or family clusters) and (2) basic molecular biology research. Finally, 34 studies meeting these criteria were used for further analyses.

We performed random effects meta-analysis to estimate the pooled risk ratio with 95% CI of 10 comorbidities for patients with severe versus non-severe COVID-19. The Mantel–Haenszel method was used to estimate the pooled effects of results [115]. The DerSimonian–Laird method was used to estimate the variance among studies [116]. Continuous data such as IL-6 levels were transformed to mean and standard deviation first using Wan’s approach based on sample size, median, and interquartile range [117]. Next, we used the inverse variance method to estimate the pooled mean difference and estimated the variance among studies using the DerSimonian–Laird method. We estimated the pooled prevalence of 3 COVID-19 symptoms (abdominal pain, diarrhea, and dyspnea) and 1 comorbidity (COPD) in 3 COVID-19 patient groups (severe, non-severe, and all). A random intercept logistic regression model was used to estimate pooled prevalence, and a maximum-likelihood estimator was used to quantify the heterogeneity of studies [118]. The tau 2 and I 2 statistics were calculated for the heterogeneity among studies. We considered I 2 ≤ 50% as low heterogeneity among studies, 50% < I 2 ≤ 75% as moderate heterogeneity, and I 2 > 75% as high heterogeneity. All meta-analyses were conducted using the meta and dmetar packages in the R v3.6.3 platform.

Building the human protein–protein interactome

A total of 18 bioinformatics and systems biology databases were assembled to build a comprehensive list of human PPIs with 5 types of experimental evidence: (1) protein complexes data identified by a robust AP-MS methodology collected from BioPlex V2.016 [119] (2) binary PPIs tested by high-throughput yeast-two-hybrid (Y2H) systems from 2 publicly available high-quality Y2H datasets [120,121] and 1 in-house dataset [32] (3) kinase–substrate interactions identified by literature-derived low-throughput or high-throughput experiments from Kinome NetworkX [122], Human Protein Resource Database (HPRD) [123], PhosphoNetworks [124], PhosphoSitePlus [125], DbPTM 3.0 [126], and Phospho.ELM [127] (4) signaling networks identified by literature-derived low-throughput experiments from SignaLink 2.0 [128] and (5) literature-curated PPIs identified by AP-MS, Y2H, literature-derived low-throughput experiments, or protein 3D structures from BioGRID [129], PINA [130], INstruct [131], MINT [132], IntAct [133], and InnateDB [109]. Inferred PPIs based on gene expression data, evolutionary analysis, and metabolic associations were excluded. Genes were mapped to their Entrez ID based on the NCBI database [134]. The official gene symbols were based on GeneCards (https://www.genecards.org/). The final human protein–protein interactome used in this study included 351,444 unique PPIs (edges or links) connecting 17,706 proteins (nodes). Detailed descriptions for building the human protein–protein interactome are provided in our previous studies [31–33,135]. An overview of the human protein–protein interactome can be found in S19 Fig.

Network proximity measure

We used the “closest” network proximity measure throughout this study. For 2 gene/protein sets A and B, their closest distance dAB was calculated as (2) where d(a,b) is the shortest distance of a and b in the human interactome. To evaluate the significance, we performed a permutation test using randomly selected proteins from the whole interactome that were representative of the 2 protein sets being evaluated in terms of their degree distributions. We then calculated the Z score as (3) where and σr were the mean and standard deviation of the permutation test. All network proximity permutation tests in this study were repeated 1,000 times.

Network-based comorbidity analysis

To reveal potential COVID-19 comorbidities, we computed the network proximity of the disease-associated proteins for each disease and the 5 SARS-CoV-2 datasets. SARS-CoV-2 target proteins with a non-negative tissue specificity in lung were used in the computation. The degree enrichment for protein i in a subnetwork was calculated as (4) where di is the degree of i in the subnetwork, n is number of nodes in the subnetwork, Di is the degree in the complete human protein interactome, and N is the total number of nodes in the interactome. The log10 ei value is reported.

We also computed the eigenvector centrality [136] of the nodes to evaluate their influence in the network topology while also considering the importance of their neighbors. A high eigenvector centrality value suggests that the node is connected to many other nodes with high eigenvector centrality scores as well. The computation was performed using Gephi 0.9.2 (https://gephi.org/).

Bulk and single-cell RNA-Seq data analysis

A list of the datasets used in this study can be found in S1 Table.

Bulk RNA-Seq datasets for asthma patients were retrieved from the NCBI GEO database (https://www.ncbi.nlm.nih.gov/geo/) using the accession numbers GSE63142 [66] and GSE130499 [67]. Differential expression of 3 comparisons—severe versus control, mild versus control, and severe versus mild—were performed using the GEO2R function (https://www.ncbi.nlm.nih.gov/geo/geo2r/) [137]. In GSE63142 [66], bronchial epithelial cells of 27 control samples, 72 mild asthma samples, and 56 severe asthma samples were obtained by bronchoscopy with endobronchial epithelial brushing. In GSE130499 [67], bronchial epithelial cells of 38 control samples, 72 mild asthma samples, and 44 severe asthma samples were available by bronchoscopy with endobronchial epithelial brushing as well. The differential expression analysis was performed by defining the groups in GEO2R first, then by selecting the 2 groups to compare. Genes with |log2FC| > 0.5 and FDR < 0.05 were considered significantly differentially expressed.

Single-cell data of normal lung and primary human bronchial epithelial cells were downloaded from https://data.mendeley.com/datasets/7r2cwbw44m/1 [14]. These datasets contain 39,778 lung cells and 17,451 bronchial epithelial cells with cell type annotated. GSE134809 [72] was downloaded from the NCBI GEO database. This dataset contains 67,050 inflamed and uninflamed cells from the ileal samples of 8 patients with Crohn disease. Qualifying cells based on the criteria from the original paper were used for the single-cell analysis. We used the cell type gene markers from a previous study [72] (CD3D, CD2, CD7, TNFRSF17, MZB1, BANK1, CD79B, CD22, MS4A1, HLA-DRB1, HLA-DQA1, LYZ, IL3RA, IRF7, GZMB, LILRA4, CLEC4C, TPSAB1, CMA1, KIT, PLVAP, VWF, LYVE1, CCL21, COL3A1, COL1A1, ACTA2, GPM6B, S100B) for the non-epithelial cells. We used markers from Zhang et al. [15] (DEFA5, REG3A, DEFA6, SOX4, CDCA7, KIAA0101, TOP2A, MKI67, HMGB2, STMN1, SPINK4, ITLN1, REG4, CLCA1, FCGBP, HMGA1, EIF3F, ETHE1, ADH1C, C1QBP, RBP2, APOB, APOC3, APOA1, APOA4) for the epithelial cells. The expression of these markers in the cells can be found in S20 and S21 Figs. All single-cell data analyses and visualizations were performed with the R package Seurat v3.1.4 [138]. “NormalizeData” was used to normalize the data. “FindIntegrationAnchors” and “IntegrateData” functions were used to integrate cells from different samples. UMAP was used as the dimension reduction method for visualization.

Building the metabolite–enzyme network

We built a comprehensive metabolite–enzyme network by assembling data from 4 commonly used metabolism databases: KEGG [139], Recon3D [140], the Human Metabolic Atlas (HMA) [141], and the Human Metabolome Database (HMDB) [142]. The metabolite–enzyme network contains 60,822 records of 6,725 reactions among 3,808 metabolites and 3,446 genes. Four types of enzyme functions were included in the network: biosynthesis, degradation, transformation, and transportation.

Building the drug–target network

To evaluate whether a drug is closely associated with SARS-CoV-2 target proteins in the human interactome, we gathered the drug–target interaction information from several databases: DrugBank database (v4.3) [143], Therapeutic Target Database (TTD) [144], PharmGKB database, ChEMBL (v20) [145], BindingDB [146], and the IUPHAR/BPS Guide to Pharmacology [147]. We included the interactions that have binding affinities Ki, Kd, IC50, or EC50 ≤ 10 μM and a unique UniProt accession number with “reviewed” status. The details for building the experimentally validated drug–target network can be found in our recent studies [31–33].

Network-based drug repurposing

We computed the closest network proximity as described before for 2,938 FDA-approved or investigational drugs and the 5 SARS-CoV-2 datasets. For prioritization, we ranked the drugs by their distance to the datasets (D < 2, network distance using the closest measure) and Z score (Z < −1.5) from the network proximity analysis. The antiviral profiles of the highlighted drugs were manually curated. COVID-19-related clinical trials were retrieved on August 28, 2020.

Gene set enrichment analysis (GSEA)

The GSEA was conducted as described in our recent work [30] as an additional source of evidence for drug repurposing. Briefly, for each drug and coronavirus target gene set, we computed an enrichment score (ES) to indicate whether the drug can reverse the effect of SARS-CoV-2 at the transcriptome or proteome level. Gene expression profiles for the drugs were retrieved from the Connectivity Map (CMAP) database [148]. Five gene sets were evaluated: (1) the DEGs in human bronchial epithelial cells infected with SARS-CoV-2 [21] (GSE147507) (2) the DEPs in human Caco-2 cells infected with SARS-CoV-2 [22] (3 and 4) 2 transcriptome datasets of SARS-CoV-1-infected samples from patient’s peripheral blood [149] (GSE1739) and Calu-3 cells [150] (GSE33267), respectively and (5) the DEGs in MERS-CoV-infected Calu-3 cells [151] (GSE122876).

The ES was calculated for up- and down-regulated genes separately first. The overall ES was calculated as (5) where ESup and ESdown were calculated using aup/down and bup/down as (6) (7) j = 1,2,⋯,s were the genes in the gene sets sorted in ascending order by their rank in the drug profiles. V(j) indicates the rank of gene j, where 1≤V(j)≤r, with r being the number of genes from the drug profile. Then, ESup/down was set to aup/down if aup/down>bup/down, and was set to −bup/down if bup/down>aup/down. A permutation test was performed to evaluate the significance. Drugs were prioritized and selected if ES > 0 and P < 0.05.

Patient data validation of the network-identified drugs using a COVID-19 registry

We used institutional-review-board-approved COVID-19 registry data, including 26,779 individuals (8,274 SARS-CoV-2 positive) tested during March 8 to July 27, 2020, from the Cleveland Clinic Health System in Ohio and Florida. The pooled nasopharyngeal and oropharyngeal swab specimens were tested. SARS-CoV-2 positivity was confirmed by reverse transcription–polymerase chain reaction assay in the Cleveland Clinic Robert J. Tomsich Pathology and Laboratory Medicine Institute. All SARS-CoV-2 testing was authorized by the FDA under an Emergency Use Authorization and complied with the guidelines established by the Centers for Disease Control and Prevention. Data include COVID-19 test results, baseline demographic information, medications, and all recorded disease conditions. We conducted a series of retrospective case–control studies with a new user active comparator design to test the drug–outcome relationships for COVID-19. Patients were actively taking the evaluated drugs (carvedilol and melatonin) at the time of testing. Data were extracted from electronic health records (EPIC Systems) and were manually checked by a study team trained on uniform sources for the study variables. We collected and managed all patient data using REDCap electronic data capture tools. The exposures of drugs (including carvedilol and melatonin) were used as recorded in the medication list in the electronic medical records at the time of testing for SARS-CoV-2. A positive laboratory test result for SARS-CoV-2 was used as the primary outcome. PS was used to match patients to reduce various confounding factors. Four models, from less to more stringent in terms of patient matching and OR adjustment, were performed: (1) model 1 was matched using age, sex, race, and smoking without adjustment for the OR (2) model 2 was matched using age, sex, race, and smoking, and the OR of COVID-19 was adjusted by age, sex, race, and smoking (3) model 3 was matched using age, sex, race, smoking, coronary artery disease, diabetes, hypertension, and COPD without adjustment for the OR and (4) model 4 was matched using age, sex, race, smoking, coronary artery disease, diabetes, hypertension, and COPD, and the OR of COVID-19 was adjusted by age, sex, race, smoking, coronary artery disease, diabetes, hypertension, and COPD. All analyses were conducted using the matchit package in the R v3.6.3 platform.

Statistical analysis and network visualization

Statistical tests were performed with the Python package SciPy v1.3.0 (https://www.scipy.org/). P < 0.05 was considered statistically significant throughout this study. Networks were visualized using Gephi 0.9.2 (https://gephi.org/).

Ethics statement

Procedures follow institutional guidelines for research on the COVID-19 registry database and were approved by the Cleveland Clinic Foundation Institutional Review Board.


Watch the video: Ενδομυϊκή Ένεση (May 2022).