Sepsis, characterised by a dysregulated host-pathogen inflammation, is a leading cause of death worldwide, and approximately 5.3 million people died from sepsis annually (1). There is a global increase of patients with distinct immunocompromised conditions due to the advance of cancer treatment, increasing biologics, and immunosuppressants for autoimmune diseases and growing organ transplant recipients (2), and it has been estimated that patients with immunocompromised conditions account for approximately 35% of all intensive care unit (ICU) admissions (3). Therefore, there is a crucial need to explore the alternation of immunological responses in sepsis among patients with immunocompromised conditions.

Immunoparalysis, characterised by not only immunologic but also by metabolic dysregulation after sepsis, has been increasingly recognised and attributed to be one of the key biological bases of prolonged impact on long-term mortality in patients with sepsis (4, 5). Recent advances in sequencing technology, such as RNA-Sequencing (RNA-Seq), immune repertoire sequencing, and single-cell RNA-Seq (sc-RNA-Seq), allow for addressing immunological and metabolic features in patients with sepsis and coronavirus disease 2019 (COVID-19) infection (6–8). Several studies have used MicroArray and RNA-Seq to identify the transcriptomic signature, so-called subendotype, in patients with sepsis, and the currently identified sepsis-associated subendotypes included early improvement of organ dysfunction after sepsis and the responsiveness of steroids in patients with a septic shock (6, 9, 10). Additionally, analytic tools, namely, MiXCR, have been recently developed to investigate the diversity of T cell receptor (TCR) using bulk RNA-Seq data (11, 12). Despite increasing studies to explore transcriptomic features of sepsis, the transcriptomic signature of sepsis among immunocompromised patients remains unclear. In the present study, we used RNA-Seq to identify distinct transcriptomic features and MiXCR to explore the diversity of TCR between immunocompromised and severity-matched immunocompetent patients with sepsis.

In the present study, we used RNA-Seq, analysed by functional pathway analyses and MiXCR, to identify immunological and metabolic features in immunocompromised patients with sepsis. We found impaired T cell-associated signalling, altered metabolic signalling, and decreased diversity of TCR among immunocompromised patients with septic. The aforementioned immunological and metabolic features indicate a dysregulated recovery in patients with septic with immunocompromised conditions, such as solid cancer underwent therapy, active haematological cancer, and receiving immunosuppressants, due to autoimmune disease or renal transplant recipient. These findings demonstrate the application of RNA-Seq to characterise the subendotype with immunoparalysis in patients with septic, particularly among immunocompromised patients.

The WHO recognises sepsis as a global health priority in 2017 (21), and it is estimated that that sepsis affects 48.9 million patients (38.9–62.9) worldwide and accounts for 19.7% (18.2–21.4) of all global deaths in 2017 (22). Immunocompromised conditions have been found to affect the outcome of sepsis substantially; however, the underlying mechanism remains elusive due to the complexity and high heterogeneity of sepsis (23, 24). The RNA-Seq has been increasingly used to address complex diseases, namely, sepsis, which disrupts the pathophysiological homeostasis result from complex, numerous, intertwined, and subcellular biological events (25). Several studies have used RNA-Seq to identify the subendotype in sepsis, and the identification of subendotype should be essential for individualised medicine in patients with sepsis (26). Antcliffe et al. recently conducted a post-hoc transcriptomic analysis among 176 patients in VANISH trial which aimed to explore the mortality impact of vasopressors and corticosteroid in patients with a septic shock and reported two transcriptomic sepsis response signatures (SRSs), such as relatively immunosuppressed signature (SRS-1) and relatively immunocompetent signature (SRS-2) (9). Notably, Antcliffe et al. found that the use of corticosteroids was associated with increased 28-day mortality in patients with septic with SRS-2, whereas the mortality was independent of the use of corticosteroids in patients with septic with SRS-1 (9).

In the present study, the upregulated T cell and Human Leukocyte Antigen (HLA)-associated pathway in immunocompetent patients with septic should be consistent with the aforementioned SRS-2 (Figure 2). The potentially harmful effect of corticosteroid among patients with septic with specific endotype was recently corroborated using the same dataset by Wong et al., whose team has developed a 100-gene-expression array, which was designed to assess adaptive immunity and glucocorticoid receptor signalling in paediatric septic shock (27). In brief, Wong et al. found that among the 44 patients with endotype A, a phenotype with relatively suppressed adaptive immunity, the mortality tended to be higher among the 26 patients receiving corticosteroid compared with 18 patients without corticosteroid (12/26, 46% vs. 4/18, 22%, p = 0.105) (27). Accordingly, the aforementioned evidence and our data demonstrated that the transcriptomic signature should have a crucial role in precision medicine among patients with sepsis.

Sepsis is a complex disease with a robust innate response and dysregulated inflammation, which in turn leads to organ dysfunction (28). In the present study, we used over-representation analysis (ORA) to identify several DEGs in immunocompetent patients, and the identified DEGs, namely, PRSS33, HDC, ALOX15, FCER1A, OLR1, and PCSK9, were mainly innate immunity and cellular damage-associated genes (7, 29–33). For example, ALOX15 and FCER1A were monocyte-associated genes, which have been found to play a pivotal role in the pathogenesis of bacterial sepsis (7, 31). Moreover, we also used CIBERSORT to estimate cellular types and found the relative abundance of monocyte in enrolled patients, and this finding was in line with the increasing data to show the crucial role of monocyte in the pathogenesis of sepsis (Supplemental Figure 1) (7, 31). Notably, the transcriptome was similar on day-1 among immunocompromised and immunocompetent patients (Figure 1A), and the findings highlight the crucial need of using paired day-1 and day-8 samples to address the dynamic transcriptome in patients with septic with high heterogeneity as shown in this study and the previous study (6). Furthermore, we employed functional pathway analysis to demonstrate the essential role of T cell immunity-relevant pathways among immunocompromised patients with sepsis. The aforementioned findings highlight that ORA-based analysis alone may not be adequate to address the complex disease with a high number of differentially expressed genes given that the top DEGs may represent the robust dysregulated immunological and cellular function (34). In contrast, functional pathway-based analyses enable us to identify the distinct T cell-associated pathways as we have shown in the present study (Figure 2).

Several transcriptomic studies have been used to address the T cell immunity in sepsis, and distinct terms were used, such as T cell exhaustion, immunoparalysis, and immune-suppressive signatures (4, 9, 26). Cheng et al. reported an in vitro model of innate immunotolerance/immunoparalysis defined by diminished production of lactate and pro-inflammatory cytokines in lipopolysaccharide (LPS) or Candida albicans-treated monocytes isolated from patients with sepsis and found not only compromised innate immunity but also impaired energy metabolism in patients with septic with immunoparalysis (4, 26). Increasing evidence has further validated the key role of metabolic abnormality in sepsis (35, 36), and our data also found impaired metabolic pathways, such as TCA cycle and cellular respiration, in immunocompromised patients with sepsis (Figure 3). Additionally, alteration in the metabolic state of T cells has been implicated with the impaired capacity of T cell expansion and effector function of T cells (37, 38). The impaired metabolic signalling as we have shown in this clinical study may hence be associated with the T cell paralysis in the immunocompromised patient with septic, although more mechanistic studies are warranted to elucidate the complex immune-metabolism interaction (4).

Increasing studies have shown the role of T cell exhaustion in patients with sepsis; however, few studies addressed the T cell diversity in patients with septic, particularly immunocompromised patients. Venet et al. conducted a pilot study to explore day-1 and day-7 diversity of TCR in 41 patients with a septic shock through using genomic DNA and multiplex PCR (39). Venet et al. reported that a reduced day-7 TCR β-chain diversity tended to be associated with an increased risk of mortality in the univariable analysis (39). Furthermore, patients who developed a nosocomial infection after the sepsis episode appeared to have a lower day-1 TCR β-chain diversity compared with those without nosocomial infection (39). Recently, there are crucial advances, such as 5′ rapid amplification of cDNA ends (RACE) and measurement of TCR in RNA-level (40). Genomic DNA was widely used to measure the diversity of TCR due to high stability and fixed copy number per cell; however, PCR-associated error result from unused segments and introns remains a major limitation of the DNA-based approach for TCR repertoire (40). In contrast, using RNA with a simplified PCR amplification strategy enables the comprehensive identification of TCR variants and is increasingly applicated (41). MiXCR is a framework that allows for analysing raw sequences of RNA-Seq to determine clonotypes of TCR. Therefore, the approach with MiXCR enables us to simultaneously conduct whole transcriptome and diversity of TCR (11). Similar to our approach, Zhigalova et al. recently demonstrated the utility of using MiXCR and RNA-Seq data to determine the dynamic diversity of TCR among tumour-infiltrating T cells treated with programmed death receptor 1 (PD-1) antibody in mouse melanoma model (42). In the present study, we used RNA-Seq to identify the impaired T cell signalling and MiXCR to reveal the decreased diversity of TCR among immunocompromised patients with septic, and the consistent findings indicate the immunocompromised condition-associated T cell paralysis in sepsis. Similar to our results in immunocompromised patients with septic, Cabrera-Perez et al., conducting qualitative and quantitative analyses of Ag-specific CD4 T cell populations in cecal ligation puncture-treated mice, found a prolonged alteration of Ag-specific T cell repertoire after sepsis despite the recovered number of T cells (43). Collectively, this evidence highlights the feasibility of using data of RNA-Seq and MiXCR to address the diversity of TCR in patients with sepsis and point out the crucial role of reduced diversity of TCR in immunocompromised patients with sepsis.

There are limitations to this study. First, the relatively small number of immunocompromised patients with septic. However, we enrolled severity-matched controls, and the highly comparable controls should allow us the identification of distinct pathways between the two groups. Second, the use of bulk RNA-Seq and future sc-RNA-Seq and functional experiments are warranted for validation at the cellular level. Third, we used RNA-Seq data to assess the diversity of TCR, and future T cell repertoire using 5′RACE should be needed to clarify the issue of TCR in patients with sepsis.

The increased number of immunocompromised patients with sepsis is a growing issue worldwide, and the RNA-Seq can be used to address the complex immunologic and metabolic features. In the present study, we enrolled immunocompromised patients with septic with severity-matched controls and conducted RNA-Seq with functional pathway and MiXCR analyses. We found that not only dysregulated innate immunity but also compromised T cell function, altered metabolic signalling, and decreased T cell diversity among immunocompromised patients with septic. These data highlight the feasibility of using RNA-Seq to characterise immunologic and metabolic features in immunocompromised patients with septic, and studies using sc-RNA-Seq and 5′RACE are warranted to elucidate the underlying cellular pathway.

The original contributions presented in the study are publicly available. This data can be found here: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE182522.

The studies involving human participants were reviewed and approved by Institutional Review Board approval of the Taichung Veterans General Hospital (CE20069B). The patients/participants provided their written informed consent to participate in this study.

P-LC, H-HC, C-YW, C-LW, T-MK, and W-CC: conceived and designed the experiments. P-LC, Y-HJ, and W-CC: acquired data. P-LC, T-HH, and W-CC: contributed materials/analysis tools. P-LC and W-CC: wrote the manuscript. All authors contributed to the article and approved the submitted version.

This study was supported in part by grants from Taichung Veterans General Hospital (TCVGH-1104402D, TCVGH-1104403D, and TCVGH-1114401D) and the Taiwanese Ministry of Science and Technology (MOST 109-2314-B-075A-005). The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.