Virus-Subtype-Specific Cellular and Humoral Immune Response to a COVID-19 mRNA Vaccine in Chronic Kidney Disease Patients and Renal Transplant Recipients

Patients with chronic kidney disease (CKD) or immunosuppression are at increased risk of severe SARS-CoV-2 infection. The vaccination of CKD patients has resulted in lower antibody concentrations and possibly reduced protection. However, little information is available on how T-cell-mediated immune response is affected in those patients and how vaccine-induced immune responses can neutralise different SARS-CoV-2 variants. Herein, we studied virus-specific humoral and cellular immune responses after two doses of mRNA-1273 (Moderna) vaccine in 42 patients suffering from CKD, small vessel vasculitis (maintenance phase), or kidney transplant recipients (KT). Serum and PBMCs from baseline and at three months after vaccination were used to determine SARS-CoV-2 S1-specific antibodies, neutralisation titers against SARS-CoV-2 WT, B1.617.2 (delta), and BA.1 (omicron) variants as well as virus-specific T-cells via IFNγ ELISpot assays. We observed a significant increase in quantitative and neutralising antibody titers against SARS-CoV-2 and significantly increased T-cell responses to SARS-CoV-2 S1 antigen after vaccination only in the CKD patients. In patients with vasculitis, neither humoral nor cellular responses were detected. In KT recipients, antibodies and virus neutralisation against WT and delta, but not against omicron BA.1, was assured. Importantly, we found no specific SARS-CoV-2 T-cell response in vasculitis and KT subjects, although unspecific T-cell activation was evident in most patients even before vaccination. While pre-dialysis CKD patients appear to mount an effective immune response for in vitro neutralisation of SARS-CoV-2, KT and vasculitis patients under immunosuppressive therapy were insufficiently protected from SARS-CoV-2 two months after the second dose of an mRNA vaccine.


Introduction
COVID-19, caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), can lead to severe and life-threatening disease with respiratory insufficiency [1,2]. Specifically, patients with chronic diseases such as obesity, diabetes, cardiovascular, and renal disease, as well as under immunosuppression are at a higher risk for an adverse outcome [2][3][4][5][6]. Thus, it is particularly important to protect these patients at risk from a severe course of SARS-CoV-2 infection. Several vaccines have been developed to protect from SARS-CoV-2 infection, among which mRNA vaccines, i.e., mRNA-1273 (Spikevax, Spike antibodies before vaccination ** BAU/mL 0 0 0 0

Blood Samples
The first blood sample was taken a few minutes after the administration of the first dose of the mRNA-1273 vaccine (100 µg, day 1) and a second blood sample was drawn three months (median 91 days, min 90 days, max 91 days) after the first vaccination during routine check-ups at nephrology outpatient clinics. Blood was collected in EDTA and serum tubes; peripheral blood mononuclear cells (PBMCs) were obtained from EDTA blood using density gradient centrifugation with Pancoll human (25 min, 1250 rpm, room temperature). PBMCs were then washed with PBS (phosphate buffered saline) two times (5 min, 1250 rpm, 4 • C). The PBMCs were stored in a freezing mix suspension (45% FCS (fetal calf serum), 45% RPMI (Roswell Park Memorial Institute) medium, and 10% DMSO (dimethyl sulfoxide)) for <1 year at −80 • C. Quantitative determination of antibody levels (S1-RBD IgG in BAU/mL, Abbott Architect) was performed using chemiluminescence microparticle immunoassay (CMIA) at both time points. Neutralisation plaque assay and T-cell activation assay were carried out as previously described [18,21,22]. To evaluate the neutralising properties of serum antibodies, neutralisation tests against the virus variants SARS-CoV-2 wildtype (WT), delta (B1.617.2), and omicron (BA.1) were performed.

Statistical Analysis
The statistical analysis was performed using LibreOffice Calc and R (Version 3.4.4). The Shapiro-Wilk test was used for testing normality. Non-parametric tests (Kruskal-Wallis test with Holm's adjustment method, Wilcoxon test, and Wilcoxon test for independent groups/Mann-Whitney U test) were used with a significance level of p < 0.05.

Results
Forty-seven participants with CKD, small vessel vasculitis (ANCA-positive with one exception), or KT recipients were included in the study with blood sampling and PBMC isolation on the day of vaccination. Of those, 43 patients received a second dose at an average interval of 35 days after the first dose. The four participants without a second vaccine dose were excluded from further analyses, as well as one patient with a positive in vitro nucleocapsid antibody result, indicative of infection prior to the second vaccination. This cohort then consisted of 18 CKD, 7 vasculitis, and 17 KT subjects and none of these patients had positive spike antibody detection prior to vaccination ( Table 1). The median renal function impairment in CKD patients was stage G3 (G2: 1 patient; G3: 14 patients; G4: 3 patients). All vasculitis patients were at maintenance state.

Immune Reaction to mRNA-1273-Quantitative Antibody Analysis
There were no measurable increases in antibodies in the three vasculitis (42.9%) and four KT (23.5%) subjects (spike antibodies < 3.0 BAU/mL) at three months after the first vaccination, respectively (see Supplementary Table S1). We found a significant increase in antibody titres within the disease groups (p < 0.05), except for the vasculitis group (p = 0.1003, Figure 1). The median antibody concentration after the second vaccination was 3572.9 BAU/mL in patients with CKD, 44.3 BAU/mL in those with vasculitis, and 27.8 BAU/mL in KT patients. Using the Kruskal-Wallis test, a significant difference in antibody levels after vaccination between the disease groups could be found (Table 1, p < 0.001). The post hoc analysis with Holm's adjustment method showed a significant difference between CKD and KT (p = 1.4 × 10 −6 ), and between CKD and vasculitis patients (p = 0.00039), which means that vasculitis and KT patients produced significantly fewer antibodies. A comparison of immune-suppressed versus non-immune-suppressed patients regardless of the underlying diagnosis is shown in Supplementary Figure S1. 0.1003, Figure 1). The median antibody concentration after the second vaccination 3572.9 BAU/mL in patients with CKD, 44.3 BAU/mL in those with vasculitis, and BAU/mL in KT patients. Using the Kruskal-Wallis test, a significant difference in antib levels after vaccination between the disease groups could be found (Table 1, p < 0. The post hoc analysis with Holm's adjustment method showed a significant differ between CKD and KT (p = 1.4 × 10 −6 ), and between CKD and vasculitis patients 0.00039), which means that vasculitis and KT patients produced significantly fewer bodies. A comparison of immune-suppressed versus non-immune-suppressed pat regardless of the underlying diagnosis is shown in Supplementary Figure S1.

Neutralisation Assays
In the group of vasculitis patients in particular, only a small proportion showed tralising antibodies (28.6% against the WT and delta variants), and there was hardly neutralisation against omicron. By contrast, in CKD patients, neutralising antibodies present against the WT and delta variants at 100%, respectively, and against omicron significantly lower level of 38.9%. In KT patients, no positive neutralising activity ag the omicron variant was detected, whereas some patients were protected against the (23.5%) and WT (52.9%) variants according to the in vitro analysis. Figure 2 show percentage of patients in each group with positive, borderline, and negative results fo NT50 value three months after the first dose of vaccination.

Neutralisation Assays
In the group of vasculitis patients in particular, only a small proportion showed neutralising antibodies (28.6% against the WT and delta variants), and there was hardly any neutralisation against omicron. By contrast, in CKD patients, neutralising antibodies were present against the WT and delta variants at 100%, respectively, and against omicron at a significantly lower level of 38.9%. In KT patients, no positive neutralising activity against the omicron variant was detected, whereas some patients were protected against the delta (23.5%) and WT (52.9%) variants according to the in vitro analysis. Figure 2 shows the percentage of patients in each group with positive, borderline, and negative results for the NT50 value three months after the first dose of vaccination.
A 50% neutralisation titer (NT50) value ≥ 1:32 was interpreted as positive, 1:32 > NT50 ≥ 1:16 as borderline, and NT50 < 1:16 as a negative result [22]. The NT50 value before vaccination was negative for all variants in all patients with the exception of two KT patients for variant BA.1 (borderline), which was also recorded after vaccination (see Supplementary Table S2). The data were not normally distributed. Using the Kruskal-Wallis test, a difference in neutralisation capacity three months after vaccination can be determined between the cohorts (WT, delta, and omicron, p < 0.001).
According to the post hoc analysis using the Wilcoxon test, significant differences in WT and delta variant neutralisation between CKD and vasculitis (p = 0.00035 and p = 0.0006), as well as KT patients (p = 2.3 × 10 −6 and p = 1.9 × 10 −5 ), were observed, which was also true for omicron neutralisation when comparing CKD and KT patients (p = 0.00058). Within the cohorts, there was no significant increase in neutralisation activity between the first vaccination (T1) and the three-month follow-up (T2) for the WT (p = 0.1814), delta (p = 0.3711), or omicron variants (p = 1) in vasculitis patients, whereas CKD patients showed a significant increase in neutralisation activity against all variants (WT and delta: p < 0.001; omicron: p = 0.0017). In KT patients, significant increases in viral neutralisation were found against the WT (p = 0.0059) and delta variants (p = 0.0143), but not against the omicron variant (p = 0.3711, Figure 3). A 50% neutralisation titer (NT50) value ≥ 1:32 was interpreted as positive, 1:32 > NT50 ≥ 1:16 as borderline, and NT50 < 1:16 as a negative result [22]. The NT50 value before vaccination was negative for all variants in all patients with the exception of two KT patients for variant BA.1 (borderline), which was also recorded after vaccination (see Supplementary Table S2). The data were not normally distributed. Using the Kruskal-Wallis test, a difference in neutralisation capacity three months after vaccination can be determined between the cohorts (WT, delta, and omicron, p < 0.001).
According to the post hoc analysis using the Wilcoxon test, significant differences in WT and delta variant neutralisation between CKD and vasculitis (p = 0.00035 and p = 0.0006), as well as KT patients (p = 2.3 × 10 −6 and p = 1.9 × 10 −5 ), were observed, which was also true for omicron neutralisation when comparing CKD and KT patients (p = 0.00058). Within the cohorts, there was no significant increase in neutralisation activity between the first vaccination (T1) and the three-month follow-up (T2) for the WT (p = 0.1814), delta (p = 0.3711), or omicron variants (p = 1) in vasculitis patients, whereas CKD patients showed a significant increase in neutralisation activity against all variants (WT and delta: p < 0.001; omicron: p = 0.0017). In KT patients, significant increases in viral neutralisation were found against the WT (p = 0.0059) and delta variants (p = 0.0143), but not against the omicron variant (p = 0.3711, Figure 3). In addition to the NT50 values, the 90% neutralisation titre (NT90) was determined, with an NT90 ≥ 1:16 interpreted as positive, 1:16 > NT90 ≥ 1:8 as borderline, and NT90 < 1:8 as negative. Supplementary Figure S2 shows the percentage of patients in each group with positive, borderline, and negative results for the NT90 value 3 months after the administration of the first vaccine dose. Supplementary Table S3 shows the NT50 and NT90 data. In addition to the NT50 values, the 90% neutralisation titre (NT90) was determined, with an NT90 ≥ 1:16 interpreted as positive, 1:16 > NT90 ≥ 1:8 as borderline, and NT90 < 1:8 as negative. Supplementary Figure S2 shows the percentage of patients in each group with positive, borderline, and negative results for the NT90 value 3 months after the administration of the first vaccine dose. Supplementary Table S3 shows the NT50 and NT90 data.

T-Cell-Mediated Immune Response
We next evaluated the effects of vaccination on the induction of SARS-CoV-2-specific T-cell responses. For this, we performed an interferon-gamma (INFγ) enzyme-linked immunospot (ELISpot) assay to investigate SARS-CoV-2 spike glycoprotein subunit 1 (S1)-specific T-cell activation. The median of spot forming units and percentages of positive T-cell response on the day of the first vaccine dose and at follow-up are shown in Table 2, where a positive T-cell response was assumed for a value of ≥10 IFNγ spots/10 6 PBMCs and a negative T-cell response for <9 IFNγ spots/10 6 PBMCs. With the exception of vasculitis patients, we observed an increase in the number of patients with an S1-specific T-cell response. Notably, the majority of patients in the vasculitis group had a positive response to S1-specific T-cells even before vaccination (five of seven patients) which declined three months later (Table 2; for a detailed graphical presentation including p-values and significances, see Figure 4 below). Table 2. T-cell response to SARS-CoV-2 S1 antigen on the day of the first vaccine dose (T1) and at three-month follow-up (T2).  Table 2. T-cell response to SARS-CoV-2 S1 antigen on the day of the first vaccine dose (T1) and at three-month follow-up (T2). We also studied general T-cell activation upon stimulation of cells with a T-cell activation cocktail, as described in the Materials and Methods section. This indicated that Tcell activation, in general, was most prominent in vasculitis patients, although there was no significant difference between the disease groups at T1 (p = 0.5241; chi-squared = 1.292) and T2 (p = 0.1682; chi-squared = 3.5651). In addition, no significant difference in pan-Tcell activation was observed within the different disease groups from T1 to T2 (vasculitis: p = 0.297; CKD: p = 0.722; and KT: p = 0.222 Figure 4).

Disease
By contrast, the specific T-cell response to S1 antigen stimulation increased significantly after vaccination in CKD patients only (p = 0.026, Figure 4).

Discussion
This study investigated the immune response to mRNA-1273 vaccination in patients with vasculitis, CKD, and KT recipients on a humoral and cellular level from a longitudi- We also studied general T-cell activation upon stimulation of cells with a T-cell activation cocktail, as described in the Section 2. This indicated that T-cell activation, in general, was most prominent in vasculitis patients, although there was no significant difference between the disease groups at T1 (p = 0.5241; chi-squared = 1.292) and T2 (p = 0.1682; chi-squared = 3.5651). In addition, no significant difference in pan-T-cell activation was observed within the different disease groups from T1 to T2 (vasculitis: p = 0.297; CKD: p = 0.722; and KT: p = 0.222 Figure 4).
By contrast, the specific T-cell response to S1 antigen stimulation increased significantly after vaccination in CKD patients only (p = 0.026, Figure 4).

Discussion
This study investigated the immune response to mRNA-1273 vaccination in patients with vasculitis, CKD, and KT recipients on a humoral and cellular level from a longitudinal perspective. The data showed that the quantitative IgG antibodies against S1-RBD increased significantly after vaccination in CKD and KT patients, but not in the vasculitis group. In vasculitis patients, there was also no significant overall increase in neutralising antibodies against the WT, delta, or omicron BA.1 variant. Although 57.1% of patients produced SARS-CoV-2-specific antibodies, only half of them (28.6%) had detectable neutralising antibodies against WT and delta. No effective neutralising capacity of antibodies against omicron was found in the majority of cases. This is important because neutralising antibodies inversely correlate to the risk of symptomatic SARS-CoV2 infection [23], meaning that with regard to the currently prevalent omicron variant, patients with fewer neutralising antibodies are less protected from symptomatic courses. Vasculitis patients received immunosuppressive therapy, as shown in Table 1, which negatively affected the immune response to vaccination. Rituximab therapy leads to reduced antibody titers after SARS-CoV-2 mRNA vaccination due to B-cell depletion (at least months after the last administration) [24][25][26]. A significantly lower number of neutralising antibodies was also found during glucocorticoid and/or antimetabolite therapy compared to immunocompetent persons [12]. Carruthers et al. found seroconversion in 73.3% of vasculitis patients with renal involvement despite rituximab therapy [27]. By contrast, we could not detect antibodies in any patient who received rituximab over the previous 6 months (but also not in one patient on cortisone/mycophenolate therapy). This may be due to B-cell depletion, which correlates with the antibody count [28]. In contrast to our study, neutralising antibodies against different virus variants and T-cell immunity were not investigated by Carruthers et al. [27]. Vaccination-induced T-cell responses were observed (despite rituximab therapy), where IFNγ production was lower upon additional immunosuppressive therapy with prednisone [26]. It was reported that patients with (eosinophilic) granulomatosis with polyangiitis, despite B-cell depletion and lack of humoral response, had a positive T-cell response to S peptides in 90.9% of cases, which was almost comparable to immune responses in healthy individuals [29]. In another study, the T-cell response was reduced in 58% of patients with rituximab therapy and remained stable for up to five weeks after the second vaccination [28]. Notably, our results showed that the T-cell response was positive for S1 peptide in 71.4% of patients. Importantly, we observed strong background T-cell activity in those subjects with autoimmune disease not linked to a specific response to SARS-CoV-2 S proteins, which limits the interpretation of such T-cell assays regarding the presence of a vaccine-specific response. While a certain percentage of patients with CKD showed a specific T-cell response to SARS-CoV-2 antigens, this was not found in subjects under immunosuppressive therapy.
In comparison to other studies, our results are derived from longitudinal observation with a follow-up at 3 months after the first vaccination (mean of two months after the second vaccine administration). This study design allowed us to make a direct comparison with the baseline, in contrast to the studies mentioned so far. The time to follow-up, however, could lead to divergent results. Carr et al. assumed a peak of antibody titres in renal predialysis patients at approx. 4 weeks after vaccination and assumed the greatest significance when measured > 4 weeks after the second dose, which is also the case in our study [30]. Comorbidities such as cardiovascular, cerebrovascular, and primary renal disease were not associated with antibody response, as recently shown by Buchwinkler et al. [9].
To generate a better immune reaction and counter the effect of immunosuppressive drugs that inhibit protection from SARS-CoV2 infections, the American College of Rheumatology advises vaccinating 4 weeks before the next rituximab application if the risk of COVID-19 is low for the patient [31]. Our results indicate that an adaptation of the vaccine to current virus variants is necessary to optimise the immune reaction and provide sufficient protection from neutralising antibodies.
The T-cell response increased a few days after the first vaccination and reached a peak 5-7 days after the second vaccination, with a decrease almost to values before the second vaccination after 80-120 days [32,33]. Thus, a more sustained T-cell response could have been missed by us in patients under immunosuppression. Our vasculitis cohort is heterogeneous (6 of 7 patients ANCA-positive, different immunosuppressive therapies) which may influence the results among this small number of participants. The T-cell reactions already before vaccination could be explained-at least partly-by cross-reactivity with other coronavirus strains, as already described in several studies [17,34].
Despite severe immunosuppression (10 patients on triple therapy and 7 on dual therapy), a significant immune response was detectable in KT patients. Neutralising antibodies increased significantly against the WT and delta variants, but seven KT patients showed no response (41.2%). Rincon-Arevalo et al. suspected that the immunosuppressive therapy on T-helper cells disturbed B and plasma cell differentiation [35]. Tacrolimus is known to inhibit signal transduction and cytokine release in T-cells, mycophenolate is associated with reduced seroconversion, whereas cortisone does not seem to have a major effect on seropositivity [36]. Previous studies reported a response to mRNA vaccination in 3 to 69% of KT patients [30,37]. Our results are significantly higher (76.5%) with a lower number of neutralising antibodies (52.9% for the WT). Immunosuppressive effects are dependent on several factors, such as age, number of immunosuppressive drugs, type of drug, time interval to transplantation, and BMI [37][38][39]. In contrast to other studies, the measurement after the second dose was later, and the patients were younger. Further, our analysis evaluates only the mRNA-1273 vaccine, which leads to higher rates of seroconversion than the BNT162b2 (Pfizer) mRNA vaccine [38]. Depending on the regimen used for immunosuppression, the response may differ in general but also with regard to the relative effects towards B-and T-cell responses.
CKD patients produced significantly more antibodies than vasculitis and KT patients. Similar results were observed in a few studies comparing only CKD and KT patients [9,36], although in our study CKD patients did not require dialysis. It is known from other vaccines that both the cellular and humoral immune response is lower in patients with CKD and decreases with the severity of renal impairment [36,40,41]. In contrast to the vasculitis and KT patients, none of the CKD patients in our study were receiving immunosuppressive drugs. In addition, we detected neutralising antibodies against the WT and delta variants in 100% of CKD patients, whereas only 38.9% had neutralising activity against omicron. In addition, we observed a significant increase in T-cell response after the second vaccination in CKD patients, indicating a good response of the humoral and cellular immune system to SARS-CoV-2 vaccination. Unfortunately, there is no comparison with the healthy general population in our study, and the threshold values for sufficient immune protection are not defined and largely differ depending on the methods and tests used.

Conclusions
In conclusion, the mRNA-1273 vaccine generates a significant humoral and cellular immune response after two doses in CKD patients not requiring dialysis. This is in sharp contrast to the results obtained in vasculitis and KT patients. KT patients show a significantly weaker immune response than CKD patients in terms of antibody concentrations and neutralising activity, which is almost absent against the omicron variant. Importantly, patients with vasculitis and after KT have hardly any detectable T-cell response to SARS-CoV-2-specific antigens although tests are often positive, which can be traced back to a high, unspecific background activity of T-cell activation. Vasculitis and KT patients are poorly protected from infection and severe disease, putting them at the front of the line for receiving immediate anti-viral treatment in case of infection [42].
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms11071756/s1, Figure S1: Comparison of immune responses in non-immune-suppressed (1) versus immune-suppressed (2) patients independent of the underlying cause of disease at the three-month follow-up: (A) quantitative anti-SARS-CoV-2 antibody levels and (B) T-cell response to S1 protein by IFNγ ELISpot. Boxplots represent the median, 25-75% quartiles, and minimum and maximum whiskers. Outliers are identified if greater/less than the third quartile ± 1.5 interquartile range. The Mann-Whitney U test was used to calculate p-values. Figure S2: NT90 (90% neutralising antibody) titers: Percentages of patients with negative, borderline, and positive NT90 are shown, taking into account the respective underlying disease (A-C) and the virus variant. Table S1: Patient quantitative antibodies in BAU/mL measured at each time point (t 1 = day of first vaccine dose; t 2 = approx. 3 months after vaccination). Table S2: List of patients' NT50 values at each time point (t 1 = day of first vaccine dose; t 2 = approx. 3 months after first vaccine dose). Table S3: Neutralisation titre according to underlying disease, taking into account the virus variant as well as the time of sampling (t 1 = day of first vaccine dose; t 2 = approx. 3 months after first vaccine dose).