A validated hypoxia-inducible factor (HIF) signature across tissue compartments predicts clinical outcome in human lung fibrosis

We previously reported that oxidative stress drives pseudohypoxic hypoxia-inducible factor (HIF) pathway activation to promote pathogenetic collagen structure-function in human lung fibrosis (Brereton et al., 2022). Here, through bioinformatic studies we investigate HIF pathway activation status in patients with idiopathic pulmonary fibrosis (IPF) and whether this has prognostic significance. Applying a well-established HIF gene expression signature, we classified publicly available datasets into HIF score-high and score-low groups across multiple tissue compartments. The HIF scores in lung tissue, bronchoalveolar lavage (BAL) and peripheral blood mononuclear cells (PBMC) were increased in IPF patients and significantly correlated with an oxidative stress signature consistent with pseudohypoxic HIF pathway activation. A high HIF score in BAL and in PBMC was a strong independent predictor of mortality in multivariate analysis. Thus, a validated HIF gene signature predicts survival across tissue compartments in IPF and merits prospective study as a non-invasive biomarker of lung fibrosis progression.


Abstract
We previously reported that oxidative stress drives pseudohypoxic hypoxia-inducible factor 2 (HIF) pathway activation to promote pathogenetic collagen structure-function in human lung 3 fibrosis (Brereton et al., 2022). Here, through bioinformatic studies we investigate HIF 4 pathway activation status in patients with idiopathic pulmonary fibrosis (IPF) and whether this 5 has prognostic significance. Applying a well-established HIF gene expression signature, we 6 classified publicly available datasets into HIF score-high and score-low groups across multiple 7 tissue compartments. The HIF scores in lung tissue, bronchoalveolar lavage (BAL) and 8 peripheral blood mononuclear cells (PBMC) were increased in IPF patients and significantly 9 correlated with an oxidative stress signature consistent with pseudohypoxic HIF pathway 10 activation. A high HIF score in BAL and in PBMC was a strong independent predictor of 11 mortality in multivariate analysis. Thus, a validated HIF gene signature predicts survival across 12 tissue compartments in IPF and merits prospective study as a non-invasive biomarker of lung 13 fibrosis progression. 14 Introduction 1 We previously reported that altered collagen nano-architecture is a defining feature of 2 idiopathic pulmonary fibrosis (IPF) that dysregulates extracellular matrix (ECM) structure-3 function to promote progressive lung fibrosis (1). We have recently extended these 4 observations, identifying that hypoxia-inducible factor (HIF) pathway activation is required to 5 promote pathologic pyridinlone collagen crosslinking and tissue stiffness by disproportionate 6 induction of collagen-modifying enzymes relative to TGFβ-induced collagen fibril synthesis 7 (2). Furthermore, we identified that this may occur via oxygen-independent (pseudohypoxic) 8 mechanisms, including a decrease in factor inhibiting HIF (FIH) activity due to oxidative stress 9 (2). Oxidative stress was increased in subpopulations of IPF fibroblasts whilst FIH activity 10 was significantly reduced in fibroblasts from patients with lung fibrosis resulting in HIF 11 activation under normoxic conditions. To further assess for HIF activity we applied a validated 12 HIF/hypoxia metagene signature, identifying that HIF activity was increased at sites of active 13 fibrogenesis in IPF tissue as well as within IPF mesenchymal cell populations. Intriguingly, in 14 a small lung mesenchymal stromal RNAseq data set increased HIF pathway activation assessed 15 via gene set variation analysis (GSVA) of this HIF signature was associated with disease 16 progression. Here we sought to further investigate the potential utility of this validated HIF 17 signature as a predictor of disease progression. 18 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

Figure 1. A validated HIF score is increased across tissue compartments in patients with
The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint The HIF score in BAL predicts mortality in patients with IPF. 1 We next investigated whether the HIF score itself had prognostic value in this BAL cohort. 2 Classifying patients into score-high and score-low groups based on unsupervised hierarchical 3 clustering (Figure 2-figure supplement 1A; Fig. 2A), we identified that the HIF score 4 significantly predicted survival in the whole IPF cohort ( Fig. 2B; hazard ratio, HR: 5.4; P = 5 2.55×10 -7 ), with comparable trends in each individual cohort (Figure 2-figure supplement 1B-6 D). A high HIF score was a strong independent predictor (HR: 4.1, P < 0.001) of mortality 7 including in multivariate analysis with the physiological Gender, Age and Physiology (GAP) 8 score that uses commonly measured clinical and physiologic variables to predict mortality in 9 IPF (9) (Fig. 2C). Notably, the HIF score alone performed similarly to the GAP index ( Fig.  10 2D), and when the HIF score and the GAP index were combined, the outcome prediction error 11 rate was reduced, indicating a significant added value (Fig. 2D). 12 To consider whether the change in HIF score might represent perturbations in cell types we 13 used annotated lung scRNA-seq data (10) as a reference to perform cell type deconvolution is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint the HIF score (blue) and Gender Age Physiology (GAP) index (green). Combining the HIF 10 score with GAP score (red) reduced prediction errors and resulted in better prediction. (E) Cell 11 deconvolution based on a signature matrix derived from the single-cell RNAseq found in 12 Reyfman et al was used to determine cell compositions in BAL samples from IPF patients with 13 low vs. high HIF score (GSE70867). *P < 0.05 by unpaired t-test. **P < 0.01. 14 15 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint The HIF score is increased in peripheral blood mononuclear cells (PBMC) from IPF 1

patients. 2
Although BAL provides direct insight into molecular events in the lung milieu, whole blood 3 and peripheral blood mononuclear cells (PBMC) have the advantage of being a less invasive 4 sampling method (12) and so we next applied the HIF score to PBMC transcriptome datasets. 5 We first applied the HIF gene expression signature (3) to a PBMC microarray dataset 6 (GSE38958) (13, 14) encompassing 45 healthy controls and 70 IPF patients, identifying that 7 the HIF score was increased in PBMC samples from IPF patients compared to health controls 8 ( Fig. 3A; P < 0.05). 9 We next calculated the HIF score in 2 PBMC microarray datasets of IPF patients with 10 available longitudinal outcome data (GSE28042, Pittsburgh, n=75; and GSE27957, Chicago, 11 n=45) (15, 16). Classifying patients into score-high and score-low groups ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint The HIF score in PBMC from IPF patients predicts mortality. 1 Finally, we investigated whether the HIF score had prognostic value in PBMC. We identified 2 that a high HIF score predicted mortality in IPF patients from the Pittsburgh cohort ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint Discussion 1 We previously reported that altered collagen fibril nanoarchitecture is a core determinant of 2 dysregulated ECM structure-function in human lung fibrosis (1), and that this dysregulation is 3 promoted by pseudohypoxic HIF pathway activation (2). Here, we extend these observations 4 to identify that HIF activity, as determined by a HIF gene signature, is increased in lung tissue, 5 BAL, and PBMC samples from patients with IPF, that this strongly correlates with an oxidative 6 stress signature consistent with pseudohypoxic HIF pathway activation, and that this is a strong 7 independent predictor of mortality. 8 We selected a well-validated 15-gene expression HIF signature (3-6) to investigate HIF 9 activity. In multiple studies this 15-gene signature has been identified to strongly correlate with 10 independent hypoxia/HIF signatures (4-6). Furthermore, it was shown to be the best performer 11 in a comprehensive study assessing the robustness of different hypoxia/HIF signatures (18). 12 Importantly, it performed consistently across multiple pre-processing analytical pipelines (18) Whilst hypoxia causes activation of the HIF pathway by inhibition of HIF prolyl 20 hydroxylases that hydroxylate HIF-α and target it for degradation via the ubiquitin proteasome 21 system, our recent study in IPF identified evidence of increased HIF activity under normoxic 22 conditions (pseudohypoxic HIF activation) as a consequence of oxidative stress. Although we 23 cannot exclude the possibility that hypoxia status contributed to HIF score in the cohorts 24 studied, the HIF score was a strong independent predictor of mortality in multivariate analysis 25 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint with the physiological Gender, Age and Physiology (GAP) score, concordant with HIF 1 pathway activation status being an independent biomarker of disease activity rather than simply 2 a reflection of hypoxia status determined by disease severity. Additionally, we identified that 3 the HIF score correlated with an oxidative stress gene expression signature across multiple 4 tissue compartments, consistent with systemic oxidative stress modulating HIF pathway 5 activity status (19). 6 Oxidative stress has been strongly implicated as an important profibrotic mechanism in the 7 lungs, with N-acetylcysteine (NAC), a precursor of the anti-oxidant glutathione, proposed as a 8 potential treatment for IPF. However, the role of NAC for IPF treatment remains controversial, 9 with the PANTHER study identifying no benefit compared to placebo (20). Whilst NAC is not 10 recommended as a treatment for IPF, meta-analyses continue to propose potential clinical 11 utility (21). Further consideration is warranted into whether patients with a high HIF-score 12 might benefit from NAC treatment. In support of the potential for stratified NAC treatment is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint Martins and colleagues were able to show an increase in monocytes upon CIH exposure (27), 1 potentially via modulating monocyte differentiation, proliferation, and survival (28, 29). 2 Furthermore, it has been recently reported (30) that monocytes from IPF patients are 3 phenotypically distinct compared to age matched controls and this was associated with an 4 increase in the level of macrophage colony-stimulating factor 1 (CSF-1) in serum. Notably, 5 CSF-1 gene transcription has previously been reported (31) to be dependent upon HIF 6 transcriptional activity. 7 In summary, we identify that a validated HIF score is increased across tissue compartments 8 in patients with IPF, that this strongly correlates with oxidative stress consistent with 9 pseudohypoxic HIF activation, and that an increase in this HIF score is a strong independent 10 predictor of mortality. Prospective validation of these findings is warranted which could inform 11 stratified approaches for the treatment of lung fibrosis. 12 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint the signature matrix of each cell type in IPF patients from single-cell gene expression profiling. 1 The imputation of bulk BAL sample used the signature matrix with S-batch correction and 2 1000 permutations for significance analysis. Student's t-test was used to evaluate the statistical 3 difference in each cell population between the two conditions. P-values were adjusted for 4 multiple testing using the Benjamini-Hochberg method. 5 6 Statistics 7 Statistical analyses were performed in GraphPad Prism v7.02 (GraphPad Software Inc., San 8 Diego, CA) unless otherwise indicated. We evaluated the correlations between oxidative stress 9 score and HIF score using Pearson's correlation. Normality of distribution was assessed using 10 the D'Agostino-Pearson normality test. Statistical analyses of single comparisons of two 11 groups utilised Student's t-test or Mann-Whitney U-test for parametric and non-parametric data, 12 respectively. Where appropriate, individual Student's t-test results were corrected for multiple 13 comparisons using the Holm-Sidak method. Results were considered significant if P < 0.05, 14 where *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

Supplementary Figures
1 Figure 1-figure supplement 1 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint Numbers below are n (%). P values, hazard ratio (HR), 95% confidence interval (CI) and 4 patient number (n) are indicated. 5 6 7 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint SM, Gottardi CJ, Budinger GRS, Misharin AV. Single-Cell Transcriptomic Analysis 48 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 12, 2022. ; https://doi.org/10.1101/2022.01.11.22269085 doi: medRxiv preprint