CSF3-Receptor Downregulation in Bone Marrow contributes to Ineffectiveness of GCSF in Advanced Liver Cirrhosis

Introduction: Cirrhosis patients exhibit cytopenia and variable response to GCSF, with benets in early cirrhosis. GCSF acts through the CSF3-receptor (CSF3R), and changes in CSF3R can affect the response. We investigated the possible mechanism for the lack of ecacy of GCSF in advanced cirrhosis Methods Cirrhotic patients (n = 127) and controls (n = 26) who underwent clinically indicated bone marrow (BM) examination were studied. BMs were assessed for RNA sequencing, qRT-PCR, and immunohistochemistry (IHC) for CSF3R and associated genes. Circulating GCSF, CSF3R, and carcinoembryonic antigen cell adhesion molecule-1 (CEACAM1) were measured. BM hematopoietic precursor cells and their alterations were examined by ow cytometry. The ndings were validated, where the GCSF was administered as for liver regeneration (n = 22) and severe neutropenia (n = 15)


Introduction
Bone marrow (BM) and liver retain an evolutionary conserved developmental relationship. BM is the potential reservoir of hematopoietic stem cell (HSC), mesenchymal stem cell (MSC) populations, responds favorably to the peripheral tissue injury and various stress factors. It provides cells and speci c molecules for tissue repair and regeneration. [1,2] G-CSF stimulates neutrophil production and HSC mobilization both at basal and stress-induced granulopoiesis. It acts through the colony-stimulating factor receptor (G-CSFR), also known as CSF3R or CD114. CSF3R is primarily expressed in BM precursor cells, i.e., HSCs. Though it is a cell surface receptor, an extracellular domain shedding could be detected in the peripheral circulation. [3] Clinical trials using G-CSF therapy have shown the survival bene t and e cacy to limit the complications in cirrhosis [4] and acute-on-chronic liver failure (ACLF) patients. [5] However, another study contradicted these ndings. It showed more signi cant complications and mortality rates in the G-CSF plus stem cell infusion group. [6] Overall, the tangible bene ts of G-CSF therapy remain variable and controversial.
Earlier, we have shown the quantitative and qualitative dysfunction of the BM stem cells in advanced cirrhosis. [7,8] Other studies showed role of BM in hepatic regenerative response. [9,10] A study from our group revealed that those with early cirrhosis and su cient BM reserve could respond to GCSF. [11] Another concern with the progressive stages of cirrhosis is severe cytopenia, including neutropenia.
Sporadically, the neutropenia could be refractory to G-CSF. [12] Considering these intriguing questions-(i) variable results to G-CSF therapy, (ii) effective response only in early cirrhosis, and (iii) at times cirrhosis related neutropenia; refractory to G-CSF led us to think whether these outcomes could be attributable to the variations {quantitative and qualitative (single nucleotide polymorphism)} in the CSF3R receptor. G-CSF's therapeutic impact is only justi ed if the receptor functioning is intact.
The present study aimed to assess the status of CSF3R in the BM precursor cells and its relationship with GCSF response.

Patients And Methods
Study Phases: The study conducted in 3 phases-1. First, we analyzed 127 cirrhosis patients and 26 controls (details in the next section) for CSF3R and followed them for six months for naturally acquired infection.
2. Second, we validated these ndings in two different clinical cohorts CSF3R receptor gene's variations for single nucleotide polymorphisms (SNPs) were analyzed in the third phase.

Phase-1 study participants and methods
Study group: The study included cirrhotic patients who underwent clinically indicated bone marrow (BM) examinations from Jan 2017 -June 2019 at the Institute of Liver and Biliary Sciences, New Delhi (n=127).
BM examination was done as per the clinical indications {(i.e., cytopenia, pyrexia of unknown origin, and dyserythropoiesis in refractory anemia in our previous study [13]}. The clinical indications across the CTP A, B, and C categories were similar [Table S1]. Thus, the study population included all the CTP classes of patients, and hence the results could be generalizable beyond the study participants. Control (n=26): Non-liver disease subjects who underwent BM examination for other than liver disease or chronic ailment served as controls, as described. [7,8] In these cases, the BM examination was done to investigate in ltrative or infective diseases. Those patients who did not have any systemic or BM pathology were considered as controls.
Institutional ethics committee approval: The Institutional Ethics Committee of the Institute of Liver and Biliary Sciences, New Delhi, India, approved this study protocol (IEC/81/MA06). Informed consent was obtained from patients to use their leftover samples and clinical data after rendering the diagnosis. The leftover bone marrow aspirate and peripheral blood, plasma, and cells were isolated and stored with the clinical annotation in the National Liver Disease Biobank.
Bone Marrow Examination: BM aspirates and biopsies were examined for various hematopoietic cell morphologies.
Bone Marrow biopsies RNA sequencing: Details given in Suppl. Methods 1-2: Brie y, total RNA was isolated from ve cases (control, Child-Turcott-Pugh Class A, and C). Total RNA was reverse transcribed, cDNAs library was prepared, and differential gene expression was analyzed in three groups-Control, Child-Turcott-Pugh Class A, and C.
Quantitative Reverse-Transcription Polymerase Chain Reaction: Details provided in suppl method-3. Brie y, total RNA was isolated, reversed transcribed into complementary DNA, and augmented by SYBR green. Transcripts were progressively expanded for gene-speci c primers, listed in Table S2. Phase-2 study groups BM and Peripheral blood plasma CSF3R at baseline were measured in two clinical cohorts. 1. One cohort (n=22) represents those cirrhotic patients who received G-CSF therapy in previous study [11] and where stored BM and peripheral blood plasma samples were available (retrospective). The response was considered if there is a reduction of CTP>1 and/or MELD>2 as described [14].
1. The second cohort was prospective, where G-CSF was given for severe neutropenia (<1000/cmm) in cirrhotic patients. We enrolled in these patients consecutively from June 2019 to June 2020 and assessed the response after 48 hours.
Phase-3 study: Sanger sequencing for CSF3R was carried out on peripheral blood samples of phase 1 and 2 study subjects (n=190) to discern the possible single nucleotide polymorphism in the CSF3R gene.
Statistical Analysis: Statistical analysis was carried out using Statistical Package for Social Sciences (SPSS, version 22.0) and MedCalc (version 20.0). Student's t-test/Mann-Whitney test was used to compare two groups based on the parametric data distribution (GSCF, CSF3R, CEACAM-1, and IHC) or non-parametric ( % cells). One-way ANOVA of variance for parametric data and Kruskal-Wallis test for non-parametric were applied for comparison amongst more than two groups. P < 0.05 was considered signi cant.

Primary attributes of study participants
During Jan 2017 -June 2019, 268 BM examinations were done in the Institute of Liver and Biliary Sciences for various clinical indications. One hundred forty-ve (145) were cirrhotic patients, and twentynine non-liver disease cases without pathology were considered controls. The rest of the cases (n=94) belonged to other diseases and hence were excluded. BM of six cirrhosis patients and three controls were suboptimal, and 12 cirrhotic patients had associated neoplastic pathology and excluded from the study.
Amongst the cirrhosis group, 28.3% patients belonged to CTP A (n=36), 42.6% to CTP B (n=54), and 29.1 % to CTP C (n=37) classes. The most frequent etiology was alcohol-associated cirrhosis (41.1%). These basic parameters amongst the CTP classes and controls succinctly summarized in Table- These results indicate the reduced CSF3R expression in the BM precursor cells.

CSF3R in Peripheral Blood
Since bone marrow is routinely not done in cirrhotic patients, we decided to test the CSF3R levels in the peripheral blood plasma. We assessed the levels of both ligand (G-CSF) and CSF3R in the BM and peripheral blood plasma. The plasma levels of G-CSF were increased [ Fig.2A], whereas the receptor levels were found to be lowered [ Fig.2B  We have earlier reported that the level of HSCs declines in advanced cirrhosis patients. [7] The lower levels of CSF3R receptors could be due to decreased BM progenitor cells. Hence, we enumerated these cells in various classes of cirrhosis patients. As expected, we found lower CD34+ cells in CTP class B and C than CTP A and controls [ Fig. 3A, B]. Further cell assessment revealed that the % population of CSF3R+ cells out of CD34+ cells also showed a declining trend with advancing cirrhosis [Fig. 3A, C].
The reduction of a stable pool of CD34+ cells could be due to higher G-CSF levels in the circulation. The mobilized HSCs lead to the formation of CFU and BFU in BM. We found that the CFU-GM and BFU-E were more pronounced in CTP A or B patients. [Fig.3 D].
This data indicates that the low CSF3R in advanced cirrhosis was related to decreased CD34 cells in BM and downregulation of CSF3R in the remaining CD34 cells in the BM. Thus the increasing lysosomal contents in CD34+ cells and increased ubiquitin in BM indirectly suggest that the CSF3R is degraded, possibly due to elevated CEACAM1.

CSF3R and Response to G-CSF in Cirrhosis
We took the stored plasma of patients (n=22) who received G-CSF treatment in an earlier study. (11) Fourteen were responsive, and eight were nonresponsive to G-CSF Table S4]. The baseline CSF3R levels were lower in BM and PB in patients who eventually did not respond to the therapy [ Fig. 6A]. Contrarily, G-CSF at baseline was higher in non-responders in comparison to those who responded. [ Fig. 6B

Discussion
The present study revealed that the G-CSF receptor (CSF3R) decline in liver cirrhosis. CSF3R becomes signi cantly downregulated in advanced cirrhosis. This reduction of CSF3R was related to hematopoietic niche alterations, lower HSCs count in BM, and reduced expression of CSF3R in remaining HSCs. It could be related to elevated CEACAM-1 in circulation as well as in the BM. On validation, it was noted that those patients who did not respond to G-CSF therapy had decreased CSF3R at baseline.
It is the rst human study exploring the lowered CSF3R in liver cirrhosis to the best of our knowledge. We argue that the lowered CSF3R in cirrhosis could cause a non-response or variable response to G-CSF therapy.
G-CSF has been a great success in the setting of chemotherapy-induced leukopenia and as a CD34+ mobilizing agent in BM transplant. Nevertheless, there have been concerns related to GCSF induced mobilization. These concerns are broadly related to HSCs niche disruption [17], underlying chronic ailment, i.e., diabetes [18], and genetic polymorphisms of the hematopoietic niche molecules (CSF3R, VCAM). [19,20] HSC niche disruption and underlying chronic ailment issues are interrelated. In advanced cirrhosis, we have shown that HSCs and their niche components are adversely affected. [7,8] Impaired glucose tolerance in diabetes can cause an inadequate response to GCSF. Likewise, another study from our group revealed that BM-MSCs have insulin resistance and impaired bioenergetics in cirrhotic patients. [21] The genetic polymorphisms of CSF3R were not identi ed in this study. It indicates a limited role of CSF3R polymorphisms and G-CSF therapy-related responses in cirrhosis patients. Thus, in advanced cirrhosis patients, primarily the disruption of stem cell niche and possibly an impaired glucose tolerance at the cellular level could induce a suboptimal response to GSCF.
[23] Ductular reaction and metaplastic hepatocytes cirrhosis can cause the excess CEACAM-1 in circulation. It works through the contact at the cell surface and can act both as a receptor/ligand. [22] CEACAM1 controls immunity and immune activation. [22,24,25] Exogenous cues CEACAM-1 in BM might control the CSF3R by a similar mechanism. The contact control explains that it was very prominent in immunohistochemistry and ELISA in CTP-B and C in BM. However, it was mildly upregulated at the BM transcription level. So, CEACAM-1 is not produced by marrow, somewhat peripherally from the diseased liver, and maybe from affected colonic epithelium in cirrhosis. It acts through contact on the cell surface. CSF3R receptor is processed in the endoplasmic reticulum and recycled back to the cell membrane. Alternatively, it is degraded in the lysosomes by ubiquitination. [3,26] Increased expression of TLR3 and ubiquitin and their inverse correlations with CSF3R; indirectly supports a possible lysosomal degradation of CSF3R.
Elevated GCSF could be harmful, especially if it is rapid and 20-50 times of controls, like in severe trauma. [27,28] Cirrhotic patients had elevated G-CSF, and it was more in advanced cirrhosis. Nevertheless, the rise was mild to moderate (2-4 times) in various stages, so it may not be that harmful in cirrhosis. Still persistently raised GSCF induces HSCs to refractory colony formation stages, and the BM pool of HSCs starts declining.
The study has some limitations. First, we did transcriptomics in a limited number of samples. However, we validated our ndings at multiple levels to support the granular change in CSF3R levels in liver cirrhosis patients. Secondly, a potential selection bias could be because we analyzed those cirrhotic patients who underwent bone marrow examination. However, as described in the methods section, the clinical indications across the CTP classes were similar, and we consecutively enrolled all cirrhotic patients with BM examinations. Hence, the bias was minimized.
Nevertheless, this study has important clinical signi cance. It shed light on a few mechanistic aspects, which have signi cant therapeutic implications. CSF3R assessment could be useful for those patients who could get bene t from G-CSF therapy. Thus, it will help the clinicians to make appropriate decisions for GCSF treatment. CS3R improvement can be achieved by humanized anti-CEACAM-1 IgG4 antibodylike MK-6018 [24,29] in the remaining HSCs of the BM of cirrhotic patients.

Declarations
Compliance with ethical standards Con ict of Interest: None Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the insti-tutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards (and/or in case humans were involved).     Relation of Baseline CSF3R with a response to GSCF and Post-transplant leukopenia and infections: Comparison of A. granulocyte colony-stimulating factor B. CSF3R levels in the bone marrow and peripheral blood plasma cirrhosis patients who responded to GSCF as regenerative therapy vs. nonresponders (mean±SD). Comparison of baseline C: GCSF and D. CSF3R in the peripheral blood plasma of cirrhosis patients who responded to GSCF treatment for neutropenia vs. non-responders (mean±SD).