INTRODUCTION

Numerous studies have shown that diabetes mellitus (DM) increases the risk of death from COVID-19 by five times [1, 2]. It is generally accepted that the high lethality from COVID-19 against the background of DM is due to the main complications of this disease: micro- and macroangiopathies, as well as heart and kidney failure. In addition, it was shown that acute respiratory viral infection increases the production of interferon gamma and increases muscle resistance to insulin, which complicates the course of DM. Insulin resistance, in turn, leads to compensatory hyperinsulinemia to maintain normal glucose levels. It should be noted that hyperinsulinemia can increase antiviral immunity due to direct stimulation of the functional activity of effector CD8+ T cells [3].

Effector CD8+ T cells play a key role in antiviral immunity at the initial stages of COVID-19 [4]. Major histocompatibility complex (MHC) class I molecules determine the efficiency of presentation of COVID-19 antigens. Immediately after entering the cell, SARS-CoV-2 induces the translation of its own proteins. Some of them enter the proteasome of the infected cell, are cleaved to peptides 8–12 aa long, and bind to class I MHC molecules. After binding, the complex consisting of the class I MHC molecule and the viral peptide is transferred from the Golgi complex to the cell surface, where it can be recognized by the CD8+ T-cell receptor of the effector T cell. In response to the interaction, the CD8+ T cell is activated and destroys the infected cell with the help of perforins and serine proteases [5].

MHC class I molecules are encoded by the HLA-A, HLA-B, and HLA-C genes, each of which can be represented in two variants (alleles). There are dozens of variants of each allele in the population, which encode MHC-I molecules with an individual ability to interact with foreign peptides. Individual combinations of MHC-I molecules significantly affect the severity of many infectious diseases [6].

Previously, we have shown that the HLA-I genotype is a significant risk factor for severe COVID-19 only in patients under the age of 60 years [6]. The aim of this study was to assess the contribution of the HLA-I genotype in patients with DM aged less than 60 years to the outcome of COVID-19.

MATERIALS AND METHODS

The study included 36 patients with type 1 and type 2 diabetes mellitus (ICD codes E10 and E11) who recovered from COVID-19. As a comparison group, an age-matched group of 186 patients who recovered from COVID-19 without a history of diabetes mellitus was selected.

The D-dimer level was assessed with an ACL TOP 700 coagulometer (IL Werfen, USA) using a D-Dimer HS 500 reagent kit (no. 0020500100). The level of C-reactive protein (CRP) was determined using the CRP_2 reagent kit (no. 06522059) with an ADVIA 1800 biochemical analyzer (Siemens, Germany).

Genomic DNA was isolated from frozen whole blood samples using the PROBA-GS-GENETIKA reagent kit (OOO NPO DNA-Technology, Russia). Libraries for sequencing exons 2 and 3 of the HLA-A, HLA-B, and HLA-C genes were prepared using the HLA-Expert reagent kit (OOO DNA-Technologies, Moscow, Russia). Sequencing was performed on a MiSeq platform (Illumina, San Diego, CA, United States) in 250-bp paired-end read mode using a MiSeq 600 cycles v3 reagent kit. The sequencing results were annotated using the IMGT/HLA v3.41.0 database of human major histocompatibility complex sequences [7].

The deviation of genotype frequency distributions from the canonical Hardy–Weinberg distribution was assessed using the Chi-square test, and the degree of differences in allele frequency between the studied groups was assessed using Fisher’s exact test. The significance of differences in the level of D-dimer and CRP was compared using the Wilcoxon test. Statistical analysis of the results was performed in the R environment.

RESULTS AND DISCUSSION

A sample of 222 patients who recovered from COVID-19 was formed, HLA genotyping was performed, and data on the level of D-dimer and C-reactive protein were analyzed. Characteristics of patients are summarized in Table 1. The comparison groups differed significantly in the outcome of the disease. For example, in the group of patients with DM, 8 deaths were recorded against 18 in the comparison group (odds ratio (OR) 2.65, p = 0.4, 95% confidence interval (CI) 0.9–7.2). This rate of deaths is two times higher than the previously obtained results for an age-matched sample of patients with DM [8]. The higher mortality may be due to the fact that our study included patients with moderate to severe COVID-19. The high incidence of obesity (OR 9.1, CI 3.8–22.3, p = 7.4e–08) and hypertension (OR 5.1, CI 2.3–12.1, p = 1.6e–05) in the group of patients with DM should be noted.

Table 1. Characteristics of the sample of patients with COVID-19
Full size table

Analysis of CRP levels showed that its blood plasma concentration during COVID-19 is significantly higher in the patients with DM (Fig. 1a). For example, in the patients without DM, the maximum level of CRP was 60 (8–117) mg/L versus 123 (62–170) mg/L in patients with DM (p < 0.01). CRP is predominantly produced by hepatocytes in response to the action of proinflammatory cytokines, mostly interleukins 6 and 1 [9, 10]. It is generally accepted that a high CRP level is characteristic of bacterial infections, since it is able to bind to bacterial wall polysaccharides and initiate the complement complex through C1q [11]. In COVID-19, CRP can not only suppress concomitant bacterial infection but also activate T-cell immunity. By binding to phosphatidylcholine on the surface of apoptotic alveocytes, CRP enhances the classical pathway of complement activation, which facilitates their uptake by macrophages bearing complement receptors CR3 and CR4 [12].

Fig. 1.
figure 1

(a) Maximum concentration of CRP in the comparison groups. (b) Maximum concentration of D-dimer in the comparison groups.

Full size image

In the group of patients with DM, an increased level of D-dimer was also detected (Fig. 1b). In patients without DM, the maximum level of D-dimer was 830 (453–1578) ng/mL, while the maximum values in patients with DM were 1591 (1182–5014) ng/mL (p < 0.01). D-dimer is a fibrin degradation product. Its high level in COVID-19 is associated with capillary microthrombosis and pulmonary vascular obstruction [13]. At the same time, it was shown that the level of D-dimer is not an independent risk factor for patients under the age of 70 years [14].

HLA genotyping of patients showed that the most frequent alleles in the group of patients without DM were HLA-A*02:01 (frequency 0.26) and HLA-A*01:01 (frequency 0.14), which is typical for the population of the Moscow region [6]. The small sample size did not allow us to identify the correlation of individual HLA-I alleles with DM (Table 2). However, the risk index for severe COVID-19 was assessed [6], which reflects the integral ability of an individual set of MHC-I molecules to present SARS-CoV-2 peptides (Fig. 2). Differences in the risk index between the patients without DM were nonsignificant (p = 0.69). However, in the group of patients with DM, the risk index differed significantly between the recovered patients and the patients with a lethal outcome (p = 0.011). At the threshold level of the risk index of 53, the sensitivity of the prediction of lethal outcome was 100%, and the specificity was 39.3% (AUC = 0.8). In the group of patients with DM and a risk index of more than 53, seven deaths compared to one death in the group with a risk index of less than 53 were registered (OR 10.2, p = 0.4, 95% CI 1.1–514). These results may indicate a significant contribution of the ability of MHC-I to present SARS-CoV-2 virus peptides to the severity of COVID-19 in patients with DM.

Table 2. Characteristics of MHC-I allele frequencies in the sample of patients with COVID-19
Full size table
Fig. 2.
figure 2

Correlation between the risk index for severe COVID-19 and the actual course of the disease in the comparison groups.

Full size image

CONCLUSIONS

1. Analysis of the level of CRP and D-dimer confirms the data on a stronger inflammatory response and microthrombosis in patients with diabetes as compared to patients with normal glycemic control.

2. Lethal outcomes of COVID-19 in patients with DM are associated with low affinity of the interaction of an individual set of MHC-I molecules with SARS-CoV-2 peptides.

3. This study demonstrates additional opportunities in assessing the risk of COVID-19 severity in patients with DM.