Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID- 19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active. Journal of Microbiology, Immunology and Infection 56 (2023) 526e536 Available online at www.sciencedirect.com ScienceDirect journal homepage: www.e-jmii.com Original Article Long term SARS-CoV-2-specific cellular immunity after COVID-19 in liver transplant recipients Maria J. Citores a,*, Aranzazu Caballero-Marcos b,c, Valentı́n Cuervas-Mons d,e, Roberto Alonso-Fernández f, Javier Graus-Morales g, Ana Arias-Milla d, Maricela Valerio f, Patricia Muñoz f, Magdalena Salcedo b,c a Laboratorio de Medicina Interna, Hospital Universitario Puerta de Hierro Majadahonda, IDIPHISA, Madrid, Spain b Hepatology and Liver Transplantation Unit, Hospital General Universitario Gregorio Marañón, Facultad de Medicina Universidad Complutense, Madrid, Spain c Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), Barcelona, Spain d Unidad de Trasplante Hepático, Servicio de Medicina Interna, Hospital Universitario Puerta de Hierro Majadahonda, IDIPHISA, Madrid, Spain e Departamento de Medicina, Facultad de Medicina, Universidad Autónoma de Madrid (UAM), Madrid, Spain f Department of Clinical Microbiology and Infectious Diseases, Hospital General Universitario Gregorio Marañón, Madrid, Spain g Department of Digestive Diseases, Hospital Ramón y Cajal, IRYCIS, Madrid, Spain Received 15 September 2022; received in revised form 9 February 2023; accepted 7 March 2023 Available online 17 March 2023 KEYWORDS Abstract Purpose: Long-term immunity after severe acute respiratory syndrome coronavirus Flow cytometry; 2 (SARS-CoV-2) infection in immunosuppressed patients is not well characterized. We aimed to Humoral immunity; explore the long-term natural immunity against SARS-CoV-2 in liver transplant (LT) recipients Liver transplantation; compared to the non-transplanted population (control group). Abbreviations: SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2; COVID-19, Coronavirus disease 2019; LT, liver transplant; S, Spike; N, Nucleocapside; M, Membrane; PBMC, Peripheral blood mononuclear cells; SD, standar deviation; TNF, tumor necrosis factor; IFN, interferon; IQR, interquartile range; DP, double positive; SOT, solid organ transplantation. * Corresponding author. Hospital Universitario Puerta de Hierro Majadahonda, Edificio de Laboratorios, Planta 3, Laboratorio de Medicina Interna, Joaquin Rodrigo 2, Majadahonda, 28222, Madrid. Spain. E-mail addresses: mariajesus.citores@salud.madrid.org (M.J. Citores), a.caballeromarcos@gmail.com (A. Caballero-Marcos), valentin. cuervasmons@salud.madrid.org (V. Cuervas-Mons), roberto.alonso@salud.madrid.org (R. Alonso-Fernández), jgraus63@gmail.com (J. Graus-Morales), anarmi85@gmail.com (A. Arias-Milla), mavami_valerio@yahoo.com.mx (M. Valerio), pacmunoz@pdi.ucm.es (P. Muñoz), magdalena.salcedo@salud.madrid.org (M. Salcedo). https://doi.org/10.1016/j.jmii.2023.03.003 1684-1182/Copyright ª 2023, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Journal of Microbiology, Immunology and Infection 56 (2023) 526e536 Methods: Fifteen LT recipients and 15 controls matched according to variables associated with Reactive T cells; disease severity were included at 12 months following the coronavirus disease 2019 (COVID-19) SARS-CoV-2 onset. Peripheral blood mononuclear cells were stimulated with peptide pools covering spike (S), nucleocapside (N), and membrane (M) proteins. Reactive CD4þ and CD8þ T cells were identified using flow cytometry, and cytokine production was evaluated in the culture super- natants using cytometric bead array. Serum anti-N and anti-S IgG antibodies were detected with chemiluminescence. Results: The percentage of patients with a positive response in both CD4þ and CD8þ T cells against each viral protein and IL2, IL10, TNF-a, and IFN-g levels was similar between LT recip- ients and controls. IFN-g levels were positively correlated with the percentage of reactive CD4þ (p Z 0.022) and CD8þ (p Z 0.043) T cells to a mixture of M þ N þ S peptide pools. The prevalence and levels of anti-N and anti-S IgG antibodies were slightly lower in the LT re- cipients, but the difference was not statistically significant. Conclusion: LT recipients exhibited a similar T cell response compared to non-transplanted in- dividuals one year after COVID-19 diagnosis. Copyright ª 2023, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). Introduction SARS-CoV-2 vaccination. LT recipients with clinical opera- tional tolerance and controls receiving any immunosup- Coronavirus disease 2019 (COVID-19) caused by severe pressive treatment were excluded. acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has All of the patients had been enrolled in previous studies5 rapidly become a public health issue profoundly affecting including a large cohort of LT recipients and controls healthcare systems worldwide, including liver trans- matched using a propensity score according to demographic plantation (LT) programs. Knowledge of the natural history features, comorbidities (diabetes, arterial hypertension, and of the disease is progressively increasing; however, the cardiovascular disease), hospital admission, requirement of durability of adaptive immunity to SARS-CoV-2 after natural mechanical ventilation, and admission to the intensive care infection in immunocompromised patients remains unclear. unit. COVID-19 was confirmed by a real-time reverse The induction of effective and durable immune memory, transcription-polymerase chain reaction assay6 on nasopha- in both humoral and cellular arms, is essential to prevent ryngeal swab samples. Severe COVID-19 was defined as severe disease and protect against reinfection. A broad admission to the intensive care unit or requirement of me- immune response to SARS-CoV-2 has been well docu- chanical ventilation. Demographic and clinical data were mented, with multiple epitopes of membrane (M) glyco- obtained from reliable electronic medical records. protein, nucleocapsid (N) phosphoprotein, and spike (S) The study was performed according to the principles of protein being the most prominent targets of specific T and B the Declaration of Helsinki and European Union Regulation cells.1,2 Previous studies have shown long-term persistence 2016/679 and was approved by the local research ethics of S-specific IgGþ memory B cells and virus-specific CD4þ committee. All patients provided written informed consent and CD8þ T cells3 in immunocompetent patients. In addi- before their inclusion in the study. tion, early and medium-term T cell-mediated immune response4 has been described in LT recipients. Compared to Sample collection nontransplanted patients, LT recipients show a lower prevalence of anti-SARS-CoV-2 IgG antibodies 12 months Peripheral venous blood was collected from each patient at after COVID-195 but the long-term cellular immunity to Hospital General Universitario Gregorio Marañón. Serum SARS-CoV-2 in these patients has not yet been assessed. samples were recovered from anticoagulant-free tubes The present study aimed to assess the long-term specific after being allowed to clot at room temperature and then SARS-CoV-2 T cell-mediated immune response in LT re- centrifuged for 10 min at 800g. The recovered serum was cipients compared with non-transplant patients. cryopreserved in small aliquots at 80  C until use. Freshly EDTA-anticoagulated blood was shipped at room tempera- ture to the Hospital Universitario Puerta de Hierro Maja- Methods dahonda, and peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation within 4 h Study population of collection. Samples shipped for cellular assays were anonymized and blinded to patient status until statistical Fifteen LT recipients and 15 non-transplanted controls who analysis was performed. had confirmed COVID-19 in MarcheApril 2020 were included at 12 months following COVID-19 diagnosis to determine In vitro stimulation of SARS-Cov-2 reactive T cells SARS-CoV-2-reactive T cells and humoral responses. None of the patients received therapy with immunoglobulins or Fresh PBMCs were stimulated with overlapping peptide convalescent plasma transfusions, active chemotherapy, or pools (PepTivator peptide pools; Miltenyi Biotec, Cologne, 527 M.J. Citores, A. Caballero-Marcos, V. Cuervas-Mons et al. Germany) covering the complete sequence of proteins M Statistical analysis and N or the immunodominant sequence domains of protein S. PBMCs were seeded at a density of 5  106 cells/cm2 in U- Continuous variables are reported as median and inter- bottom 96-well plates in TexMACS (Miltenyi Biotec) cul- quartile range (IQR), and categorical variables as absolute ture medium containing 1ug/ml of proteins M, N, S, or a numbers and percentages. mixture of all three proteins (M þ N þ S). Each experiment Demographic and clinical characteristics of LT recipients included negative and positive controls for stimulation, and controls or of patients classified according to the consisting of PBMCs in complete medium alone or with a reactivity of T cells against SARS-CoV-2 protein N were polyclonal stimulator of human effector/memory T cells compared using the Fisher correction for the chi-square (Cytostim; Miltenyi Biotec), respectively. Cells were har- test or the ManneWhitney U test, whenever appropriate. vested after 16 h at 37  C and 5% CO2, and culture super- The ManneWhitney U test was used to compare the natants were cryopreserved at 20  C until use. percentage of reactive T cells and cytokine or IgG levels between the LT recipients and controls. Qualitative ana- lyses of reactive T-cells against SARS-CoV-2 proteins and IgG Analysis of SARS-CoV-2- reactive T cells prevalence were performed using Fisher’s correction for the chi-square test. Correlations between the percentage of reactive T cells After culturing under the conditions described above, PBMCs and cytokine or serum IgG levels were calculated using were washed in PBS/2 mM EDTA/0.5% BSA and incubated with Spearman’s test. mouse anti-human antibodies anti-CD3-FITC (clone BW264/ Statistical analyses were made using SPSS version 25 (IBM 56), anti-CD137-PE (clone 4B4-1), anti-CD69-APC (clone SPSS Statistics), and graphs were generated using GraphPad FN50, all from Miltenyi Biotec), and anti-CD4-PerCp-Cy5.5 or Prism version 6.0 (GraphPad Software). All tested hypoth- anti-CD8-PerCp-Cy5.5 (BD Biosciences; Franklin Lakes, New eses were two-tailed and considered significant at P < 0.05. Jersey, USA). Mouse anti-human isotype control-irrelevant antibodies were used as negative controls for staining. At least 30,000 CD3þCD4þ or CD3þCD8þ cells were acquired Results and analyzed using a FACSCalibur flow cytometer (BD Bio- sciences) with CELLQuest-Pro software. SARS-CoV-2 reactive T cells were considered as Study population and baseline characteristics CD3þCD4þ or CD3þCD8þ T lymphocytes expressing CD137 and CD69 simultaneously. The percentage of reactive T The assessment of T cell-mediated and humoral immune cells was calculated by subtracting non-specific reactive T responses was carried out at a median time of 11.9 (IQR cells when PBMCs were cultured in media alone. The cut-off 11.7e12.16) months after COVID-19 diagnosis without evi- values for a positive response were set at the mean  2 dence of SARS-CoV-2 reinfection. Most patients were male standard deviation (SD) of reactive T cells when PBMCs (73.3%), with a median age of 63 years (Table 1). Arterial were cultured without stimuli. hypertension was the most frequent comorbidity (66.7%), and none of the patients had chronic obstructive pulmonary disease. All patients presented with symptomatic COVID-19. Eleven (73.3%) LT recipients and 14 (93.3%) controls required Analysis of cytokine secretion after stimulation hospital admission, but only 4 patients (13.3%) presented with SARS-CoV-2 peptides with severe COVID-19. None of the LT recipients received interferon beta, and they were less frequently treated with Interleukin (IL)-2, IL-4, IL-6, IL-10, tumor necrosis factor lopinavir (p Z 0.005) than were controls (Table 1). At the (TNF)-a, and interferon (IFN)-g were quantified in the cul- time of COVID-19 diagnosis, all LT recipients received chronic ture supernatants by flow cytometry using the Cytometric immunosuppression, mainly tacrolimus (n Z 11; 73.3%), Bead Array Human Th1/Th2 Cytokine Kit II (BD Biosciences) followed by mycophenolate mofetil (n Z 7; 46.7%) and according to the manufacturer’s instructions. Samples and everolimus (n Z 4; 26.7%). Twelve months post-infection, standards were acquired using a FACSCalibur flow cytome- the majority of patients received tacrolimus (n Z 11; ter and analyzed using FCAP Array software (BD Bio- 73.3%) and everolimus (n Z 6; 40.0%) (Table 1). sciences). The final concentration of each cytokine was calculated by subtracting the concentration observed in the SARS-Cov-2 reactive T cells supernatant of cells cultured in the media alone. The percentage of SARS-CoV-2 reactive T cells against viral proteins was calculated in gated CD4þ and CD8þ T cells Assessment of SARS-CoV-2-specific antibodies (Fig. 1). The percentage of SARS-CoV-2 reactive T cells in PBMCs cultured in media alone was 0.02  0.014% and SARS-CoV-2 IgG antibodies targeting protein N were 0.019  0.014% in CD4þ and CD8þ T cells, respectively, and detected in serum samples with chemiluminescence (SARS- the cut-off value for a positive response was set at 0.048% CoV-2 IgG Reagent Kit, Abbott, Chicago, Illinois, USA)5 and for CD4þ T cells and 0.047% for CD8þ T cells. antibodies targeting protein S with a quantitative chemi- No differences were found in the percentage of reactive luminescent assay (SARS-CoV-2 IgG II Quant Reagent Kit, CD4þ (31.1  13.9% and 24.3  7%; p Z 0.172) or CD8þ Abbott)5 using the ARCHITECT i2000 INSTRUMENT (Abbott). (35.1  12.8% and 37.7  13.3%; p Z 0.847) T cells between 528 Journal of Microbiology, Immunology and Infection 56 (2023) 526e536 Table 1 Demographical and clinical characteristics of the study group. LT recipients (n Z 15) Immunocompetent controls (n Z 15) p Age (years) 63 (43e66) 62 (47e72) 0.461 Sex (male) 11 (73.3) 11 (73.3) >0.99 Previous medical history Diabetes mellitus 5 (33.3) 7 (46.7) 0.710 Hypertension 9 (60.0) 11 (73.3) 0.700 ACE inhibitors or ARB 6 (40.0) 8 (53.3) 0.715 Clinical characteristics Non-severe COVID-19a 14 (93.3) 13 (86.7) >0.99 Hospital admission 11 (73.3) 14 (93.3) 0.330 Interval since transplantation (years) 5.57 (2.14e12) NA NA COVID-19 specific therapy Lopinavir 6 (40.0) 14 (93.3) 0.005 Interferon beta 0 10 (66.7) 0.001 Hydroxychloroquine 13 (86.7) 14 (93.3) >0.99 Azithromycin 9 (60.0) 4 (26.7) 0.114 Tocilizumab 1 (6.7) 1 (6.7) >0.99 Corticosteroids (boluses) 1 (6.7) 1 (6.7) >0.99 Immunosuppression at baseline Tacrolimus 11 (73.3) NA NA Cyclosporine 1 (6.7) NA NA Mycophenolate 7 (46.7) NA NA Corticosteroids (maintenance) 1 (6.7) NA NA Everolimus 4 (26.7) NA NA Immunosuppression at month 12 Tacrolimus 11 (73.3) NA NA Tacrolimus trough levels (ng/mL), n Z 11 3.7 (1.9e4.7) NA NA Cyclosporine 0 NA NA Mycophenolate 3 (20.0) NA NA Corticosteroids (maintenance) 0 NA NA Everolimus 6 (40.0) NA NA Everolimus trough levels (ng/mL), n Z 6 3.25 (2.9e4.9) NA NA a Severe COVID-19 was defined as the requirement for respiratory support or admission to the intensive care unit. Data are expressed as the median (interquartile range) or n (%). ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; NA, not applicable. LT recipients and controls, respectively, when PBMC were lowest proportion of positive responses in both CD4þ and cultured with the positive control of antigenic stimulation. CD8þ T cells (Table 2). All LT recipients and controls presented SARS-CoV-2 Two LT recipients and one control failed to show a pos- reactive CD4þ and CD8þ T cells when stimulated with a itive response to protein N in both CD4þ and CD8þ T-cells. mixture of M þ N þ S proteins, and no differences between To explore potential risk factors for a lower or null response the two groups were found in the percentage of reactive T to protein N, we compared demographic and clinical data cells (Fig. 1). When analyzing the reactive T cells against of patients with both reactive CD4þ and CD8þ T cells each viral protein separately, the percentage of reactive [double positive (DP)] response (eight LT recipients and CD8þ T cells was similar for the two study groups. However, seven controls) and patients who did not show a DP LT recipients showed a higher percentage of CD4þ T cells response to protein N. All variables considered were similar reactive against S (p Z 0.003) and M (p Z 0.037) proteins for patients with and those without a DP response (Table 3), than did the control group (Fig. 1). although LT recipients receiving tacrolimus tended to pre- Considering reactive T cells in a qualitative trait, no sent a non-DP response to protein N (p Z 0.077). differences were found between the percentage of LT recipients and controls with a positive response to each Cytokine production after stimulation with SARS- viral protein (Table 2). All patients showed a positive Cov-2 peptides CD4þ and/or CD8þ T-cell response to at least one of the viral proteins analyzed. In addition, 15 patients (eight LT Interleukine-6 was widely detected above the maximum recipients and seven controls) showed a positive response quantifiable level with the kit used, while IL4 concentration of both CD4þ and CD8þ T cells to all three proteins. A was below detectable levels in all culture conditions. higher proportion of positive responses was found in CD4þ Therefore, IL6 and IL4 production after stimulation with T cells than in CD8þ T cells, and protein N induced the SARS-Cov-2 peptides was excluded from further analysis. 529 M.J. Citores, A. Caballero-Marcos, V. Cuervas-Mons et al. Figure 1. Reactive CD4D and CD8D T cells against SARS-CoV-2 proteins. PBMC from 15 LT recipients and 15 controls were stimulated in vitro with proteins M, N, S, and a mixture of M þ N þ S for 16 h (A) Lymphocytes were gated according to FSC-height and SSC-height characteristics. Then, CD3þCD4þ or CD3þCD8þ T cells were selected, and SARS-CoV-2 reactive T cells were considered as those CD69þCD137þ cells in gated CD3þCD4þ or CD3þCD8þ cells (B) Example plots showing staining patterns of SARS-CoV-2 reactive T cells in gated CD3þCD4þ and CD3þCD8þ T cells in each culture condition (C) Percentage of reactive CD4þ and CD8þ T cells against each SARS-CoV-2 protein in 15 LT recipients and 15 controls. Bars show median percentage and error bars show interquartile range. Under almost all culture conditions, IL2, IL10, TNF-a, and IFN-g (p Z 0.049) than did patients without a DP and IFN-g levels were slightly lower in LT recipients than in response (Fig. 4). controls, although the differences were not statistically significant (Fig. 2). We then explored the correlation between cytokine Prevalence and quantitative assessment of IgG levels and the percentage of CD4þ, CD8þ, or the sum of antibodies against SARS-CoV-2 CD4þ and CD8þ (total) reactive T cells in all the included patients (n Z 30). After stimulation with M þ N þ S All patients, except for two LT recipients, showed anti-S IgG proteins, a positive correlation was found between IFN-g antibodies at 12 months following COVID-19 diagnosis. In and the percentage of CD4þ, CD8þ, and total reactive T contrast, only five (33.3%) LT recipients and eight (53.3%) cells (Fig. 3). No correlation was found between cytokine controls presented anti-N IgG antibodies. No statistical levels and the percentage of reactive T cells to proteins M significance was reached when comparing the prevalence or S. Patients with a DP response to protein N presented or levels of anti-S and anti-N IgG antibodies of the LT re- higher production of IL2 (p Z 0.033), TNF-a (p Z 0.008), cipients and controls (Table 4). 530 Journal of Microbiology, Immunology and Infection 56 (2023) 526e536 Table 2 Detectable T cells reactive to SARS-CoV-2 proteins M, N and S. All patients (n Z 30) LT recipients (n Z 15) Immunocompetent controls (n Z 15) p Reactive CD3DCD4D T cells Against protein M 29 (96.7) 15 (100) 14 (93.3) >0.99 Against protein N 25 (83.3) 12 (80) 13 (86.7) >0.99 Against protein S 30 (100) 15 (100) 15 (100) >0.99 Reactive CD3DCD8D T cells Against protein M 28 (93.3) 13 (86.7) 15 (100) 0.483 Against protein N 17 (56.7) 9 (60) 8 (53.3) >0.99 Against protein S 23 (76.7) 12 (80) 11 (73.3) >0.99 Data are expressed as n (%). Table 3 Demographical and clinical characteristics of the study group according to T cell response to SAR-Cov-2 protein N. DP response (n Z 15) No DP response p (n Z 15) Age (years) 63 (44e69) 63 (43e66) 0.539 Sex (male) 13 (86.7) 9 (60.0) 0.215 Previous medical history Diabetes mellitus 7 (46.7) 5 (33.3) 0.710 Hypertension 9 (60.0) 11 (73.3) 0.700 ACE inhibitors or ARB 5 (33.3) 8 (60.0) 0.272 Clinical characteristics Non-severe COVID-19a 14 (93.3) 12 (80.0) 0.598 Hospital admission 13 (86.7) 12 (80.0) p > 0.99 COVID-19 specific therapy Lopinavir 8 (53.3) 12 (80.0) 0.245 Interferon beta 4 (26.7) 6 (40) 0.700 Hydroxychloroquine 13 (86.7) 14 (93.3) p > 0.99 Azithromycin 5 (33.3) 8 (53.3) 0.462 Tocilizumab 1 (6.7) 1 (6.7) p > 0.99 Corticosteroids (boluses) 0 (0.0) 2 (13.3) 0.483 Immunosuppression at baseline (n Z 8) (n Z 7) Tacrolimus 4 (50) 7 (100) 0.077 Cyclosporine 1 (12.5) 0 (0.0) p > 0.99 Mycophenolate 4 (50.0) 3 (42.9) p > 0.99 Corticosteroids (maintenance) 1 (12.5) 0 (0.0) p > 0.99 Everolimus 3 (37.5) 1 (14.3) 0.569 Immunosuppression at month 12 (n Z 8) (n Z 7) Tacrolimus 4 (40.0) 7 (100) 0.077 Tacrolimus trough levels (ng/mL), n Z 11 3.6 (2.1e6.6) 4.2 (1.9e4.7) p > 0.99 Mycophenolate 2 (25.0) 1 (14.3) p > 0.99 Everolimus 3 (37.5) 3 (42.9) p > 0.99 Everolimus trough levels (ng/mL), n Z 6 3.3 (2.1e6.6) 3.2 (2.7e3.2) p > 0.99 Interval since transplantation (years) 6.9 (3.2e18.9) 2.4 (1.7e8.4) p Z 0.19 a Severe COVID-19 was defined as the requirement for respiratory support or admission to the intensive care unit. Data are expressed as the median (interquartile range) or n (%). DP: patients with both CD4þ and CD8þ positive response to protein N. Correlation between humoral and cellular response to protein S or N, respectively, except for the two LT re- to SARS-CoV-2 cipients and one control who failed to present T cells reactive against protein N. No correlation was observed between the percentage of reactive T cells against protein N or S and levels of anti-N or SARS-CoV-2 reinfections beyond one year anti-S IgG antibodies, respectively (Fig. 5), in all included patients (n Z 30). Remarkably, all patients lacking anti-S Six patients (3 LT recipients and 3 controls) documented IgG or anti-N IgG antibodies showed CD4þ T cells reactive SARS-CoV-2 reinfection, all with asymptomatic or mild 531 M.J. Citores, A. Caballero-Marcos, V. Cuervas-Mons et al. Figure 2. Cytokine production after stimulation with SARS-CoV-2 proteins. Concentration (pg/ml) of IL2, IL10, TNF-a, and IFN- g in culture supernatants after stimulation with proteins M, N, S, and M þ N þ S in 15 controls and 15 LT recipients. Bars show median percentage and error bars show interquartile range. disease, at 11.2  1.4 months following the study. All pa- N, and S), and 50% of patients showed both CD4þ and CD8þ tients had received mRNA COVID-19 vaccines after the T cells reactive against all three viral proteins. study, and reinfections were diagnosed at 5.7  2.3 months As reported in several studies, most individuals that have after two (n Z 1) or 3 (n Z 1) doses in LT recipients and at recovered from COVID-19 present T-cell responses to pro- 5.03  3.02 months after one, two, or three doses in teins M, N, and S soon after diagnosis,7 with a stable SARS- controls. CoV-2 T-cell repertoire of up to 3e8 months after No differences in cellular or humoral responses were symptoms4,10e13 regardless of disease severity. Very few found in patients and controls with or without SARS-CoV-2 studies have analyzed the T-cell response to SARS-CoV-2 one reinfections. year after infection. Zhang3 et al. found a T-cell response maintained over time in most patients at 6 and 12 months after disease onset. In a recent study performed in conva- lescent patients over a period of 1e11 months after COVID,14 Discussion specific response (T cell and IgG levels) was present in 90% of patients, but in contrast to our results, exposure to immu- The limited data available indicate that the SARS-CoV-2 nosuppressive drugs was an independent risk factor for the reactive cellular response in transplant recipients and in the absence of a specific T-cell response. However, it is impor- general population is quite similar from soon after symptom tant to note that only 2 out of the 6 patients who were onset7,8 up to six-eight months later.4,9 Our results show that receiving immunosuppressive therapy were solid organ the magnitude and functionality of cellular T-cell responses transplantation (SOT) recipients14; therefore, these results against SARS-CoV-2 were similar in LT recipients and non- in the SOT population should be interpreted with caution. transplant patients 12 months following COVID-19 infec- The percentage of reactive T cells for each viral protein tion. Remarkably, all LT recipients and controls presented T tested was highly variable, and SARS-Cov-2 reactive CD4þ T cells reactive against at least one viral protein analyzed (M, cells exceeded CD8þ T cells against each protein, as widely 532 Journal of Microbiology, Immunology and Infection 56 (2023) 526e536 Figure 3. Correlation among cytokine levels and reactive T cells against SARS-CoV-2 proteins. Correlations between levels (pg/ml) of IL2, IL10, TNF-a, and IFN-g in culture supernatants and reactive CD4þ, CD8þ and total T cells after PBMC stimulation with proteins M þ N þ S. Median percentage of reactive T cells and IgG levels of all 30 patients included are shown. described,1,4,7,11,12 in both LT recipients and controls. In quantified cytokine levels in culture supernatants rather addition, proteins M and S induced a higher percentage of than directly analyzing their production by reactive T cells reactive T cells than did protein N in both CD4þ and CD8þ T (such as intracellular staining or ELISpot), we suggest that cells.9,15 While most patients presented a positive response these cytokines may be specifically secreted by T cells of both CD4þ and CD8þ T cells against proteins M and S, reactive to SARS-CoV-2 because of the positive correlation only 50% of patients presented a DP response to protein N. between the percentage of reactive T cells and IFN-g However, no differences were found between patients with levels. Moreover, the culture supernatants of patients with and those without a DP response against protein N with a DP response of both CD4þ and CD8þ T cells against pro- respect to demographic and clinical characteristics, COVID- tein N presented higher levels of IL2, TNF-a, and IFN-g than 19 severity, or time elapsed since transplantation or did those of patients without a DP response. immunosuppressive regimen in LT recipients. No significant differences were found between LT re- In line with previous reports,1,3,7,12,15 we found polari- cipients and nontransplant patients in terms of specific zation of reactive T cells towards a classic Th1 type, as humoral immunity. However, we cannot conclude that hu- considerable IL2, TNF-a, and IFN-g were produced, while moral immunity is similar in the two groups since lower very little or no IL4 production was observed. Although we prevalence and levels of anti-S and anti-N were found in LT 533 M.J. Citores, A. Caballero-Marcos, V. Cuervas-Mons et al. Figure 4. Cytokine production in patients with and without a DP response to SARS-Cov-2 protein N. Concentration (pg/ml) of IL2, IL10, IFN-g (left), and TNF-a (right) in culture supernatants after stimulation with protein N in patients with and without a positive response to both CD4þ and CD8þ T cells (DP). Bars show median percentage and error bars show interquartile range for all 30 patients included. Table 4 Incidence and levels of anti-spike and anti-nucleocapsid IgG antibodies according to study group. LT recipients (n Z 15) Immunocompetent patients (n Z 15) p Anti-spike IgG detected; n (%) 13 (86.7) 15 (100.0) 0.483 Anti-spike IgG levels (UA/mL); median (IQR) 502.3 (169.5e984.2) 819.1 (480.9e2159.7) 0.370 Anti-nucleocapsid IgG detected; n (%) 5 (33.3) 8 (53.3) 0.462 Anti-nucleocapsid IgG levels; median (IQR) 0.8 (0.11e3.57) 1.75 (0.53e3.14) 0.232 Data are expressed as median (IQR) or n (%). recipients, although statistical significance was not the highly mutated variant Omicron21 in most individuals. In reached, perhaps due to the small sample size. this regard, none of unvaccinated patients included in this The proportion of patients with anti-S IgG antibodies one study suffered documented reinfection within one year. year after COVID-19 was very high in both study groups. In Indeed, reinfections were diagnosed about two years after contrast, a lower prevalence of anti-N IgG antibodies was prior infection and were probably due to Omicron since it found, which is in agreement with previous studies.16,17 We was the predominant strain in Spain in this period. All re- did not find a correlation between the percentage of T cells infections presented with mild diseases, but since all pa- reactive against N or S and anti-N or anti-S IgG antibody tients were vaccinated after the study, we cannot discern levels, as has also been reported in both SOT recipients and whether protection from severe disease was attributable to in the general population at 6 months following infec- natural immunity, vaccination, or both. tion.4,12,13,18 Remarkably, the vast majority of patients This study has several limitations that should be lacking anti-N IgG antibodies presented CD4þ T cells reac- considered when interpreting the results. First, the tive against N. relatively small number of patients may have limited the These results show that most LT recipients and non- ability to detect clinical associations with cellular transplant patients who recovered from COVID-19 have response. Second, a high proportion of the patients SARS-CoV-2 specific CD4þ T cells, which, along with the required hospitalization but presented a non-severe dis- reported memory B-cell repertoire,11,12,19 likely contribute ease. Therefore, it is possible that the spectrum of mild to the development of protective immunity for at least 12 and asymptomatic COVID-19 as well as the most severe months after SARS-CoV-2 infection. Therefore, serodiag- disease are not adequately captured, and our results nostic tests alone cannot be considered strong indicators of could not be extrapolated to all forms of the disease. In protective immunity in COVID-19 convalescent patients, addition, corticosteroid boluses were rarely administered and T-cell assays might be considered to investigate the to patients included in our study, although their use is history of prior infection in epidemiological studies. now widespread in the acute phase of COVID-19. Although Natural SARS-CoV-2 infection provides substantial and it is not yet clear whether corticosteroid boluses in the persistent immunologic protection to similar strains for a acute phase impair T-cell immunity in the long term, this period of several months,20 and it is also cross-reactive with possibility should be considered when interpreting our 534 Journal of Microbiology, Immunology and Infection 56 (2023) 526e536 Figure 5. Correlation among reactive T cells and anti-IgG levels against SARS-CoV-2 at 12 months after COVID-19 diagnosis. Correlation between (A) reactive T cells against protein S and anti-S IgG levels, and (B) reactive T cells against protein N and anti-N IgG levels. results. Furthermore, given the long interval since Fundacion Lair” and Aránzazu Caballero-Marcos from the transplantation of the patients included in our study, our Instituto de Salud Carlos III (Rio Hortega-CM19/00247). results should be interpreted with caution in regards to the first year post-transplantation; a higher degree of Data statement immunosuppression during the early post-transplant period theoretically could have led to differences in im- Data supporting the findings of this study are available from munity against SARS-CoV-2 between transplant recipients the corresponding author upon reasonable request. and non-transplanted patients. Finally, since all patients received SARS-CoV-2 vaccination after the study, it is not possible to determine how long the protection of natural Declaration of competing interest immunity lasts. However, long-term data regarding the T- cell immunity of unvaccinated patients are scarce; None. therefore, our results may be of interest for strategies against future SARS-CoV-2 variants that could be less covered by currently available vaccines. References In conclusion, despite exposure to immunosuppression, LT recipients exhibited similar functional T-cell responses 1. Grifoni A, Weiskopf D, Ramirez SI, Mateus J, Dan JM, against SARS-CoV-2 to the reponses of matched non- Moderbacher CR, et al. Targets of T cell responses to SARS-CoV- 2 coronavirus in Humans with COVID-19 disease and unexposed transplanted individuals one year after a COVID-19 individuals. Cell 2020;181. 1489.e15-1501.e15. diagnosis. 2. Peng Y, Mentzer AJ, Liu G, Yao X, Yin Z, Dong D, et al. Broad and strong memory CD4þ and CD8þ T cells induced by SARS- Funding CoV-2 in UK convalescent individuals following COVID-19. Nat Immunol 2020;2:1336e45. 3. Zhang J, Lin H, Ye B, Zhao M, Zhan J, Dong S, et al. One-year This study has been supported by the grant “Beca de sustained cellular and humoral immunities of COVID-19 con- Investigación de la Fundación Sociedad Española de Tras- valescents. Clin Infect Dis 2021:ciab884. plante Hepático (FSETH) 2020”. Maria J. Citores has 4. Fernández-Ruiz M, Olea B, Almendro-Vázquez P, Giménez E, received support from the “Catedra de Patrocinio UAM- Marcacuzco A, San Juan R, et al. T cell-mediated response to 535 M.J. Citores, A. Caballero-Marcos, V. Cuervas-Mons et al. SARS-CoV-2 in liver transplant recipients with prior COVID-19. 13. Kirchner T, Heinrich S, Bonifacius A, Engel B, Ruhl L, Pink I, Am J Transplant 2021;21:2785e94. et al. Reduced humoral but stable cellular SARS-CoV-2-specific 5. Caballero-Marcos A, Salcedo M, Alonso-Fernández R, Rodrı́- immunity in liver transplant recipients in the first year after guez-Perálvarez M, Olmedo M, Graus Morales J, et al. Changes COVID-19. PLoS One 2022;7:e0276929. in humoral immune response after SARS-CoV-2 infection in liver 14. Konik M, Lindemann M, Zettler M, Meller L, Dolff S, Rebmann V, transplant recipients compared to immunocompetent patients. et al. Long-term SARS-CoV-2 specific immunity is affected by Am J Transplant 2021;21:2876e84. the severity of initial COVID-19 and patient age. J Clin Med 6. Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, et al. Clinical 2021;10:4606. features of patients infected with 2019 novel coronavirus in 15. Thieme CJ, Anft M, Paniskaki K, Blazquez-Navarro A, Wuhan, China. Lancet 2020;395:497e506. Doevelaar A, Seibert FS, et al. Robust T cell response toward 7. Thieme CJ, Anft M, Paniskaki K, Blazquez-Navarro A, spike, membrane, and nucleocapsid SARS-CoV-2 proteins is not Doevelaar A, Seibert FS, et al. The magnitude and functionality associated with recovery in critical COVID-19 patients. Cell Rep of SARS-CoV-2 reactive cellular and humoral immunity in trans- Med 2020;1:100092. plant population is similar to the general population despite 16. Fenwick C, Croxatto A, Coste AT, Pojer F, André C, Pellaton C, immunosuppression. Transplantation 2021;105:2156e64. et al. Changes in SARS-CoV-2 antibody responses impact the 8. Candon S, Guerrot D, Drouot L, Lemoine M, Lebourg L, Hanoy M, estimates of infections in population-based seroprevalence et al. T cell and antibody responses to SARS-CoV-2: experience studies. J Virol 2021;95. e01828-20. from a French transplantation and hemodialysis center during the 17. Chia WN, Tan CW, Foo R, Kang AEZ, Peng Y, Sivalingam V, et al. COVID-19 pandemic. Am J Transplant 2021;21:854e63. Serological differentiation between COVID-19 and SARS in- 9. Favà A, Donadeu L, Jouve T, Gonzalez-Costello J, Lladó L, fections. Emerg Microb Infect 2020;9:1497e505. Santana C, et al. A comprehensive assessment of long-term 18. Schiffner J, Backhaus I, Rimmele J, Schulz S, Möhlenkamp T, SARS-CoV-2-specific adaptive immune memory in convales- Klemens JM, et al. Long-term course of humoral and cellular cent COVID-19 solid organ transplant recipients. Kidney Int immune responses in outpatients after SARS-CoV-2 infection. 2022;101:1027e38. Front Public Health 2021;9:732787. 10. Bonifacius A, Tischer-Zimmermann S, Dragon AC, Gussarow D, 19. Moore T, Hossain R, Doores KJ, Shankar-Hari M, Fear DJ. SARS- Vogel A, Krettek U, et al. COVID-19 immune signatures reveal CoV-2-specific memory B cell responses are maintained after stable antiviral T cell function despite declining humoral re- recovery from natural infection and postvaccination. Viral sponses. Immunity 2021;54:340e354.e6. Immunol 2022;35:425e36. 11. Sherina N, Piralla A, Du L, Wan H, Kumagai-Braesch M, 20. Spicer KB, Glick C, Cavanaugh AM, Thoroughman D. Protective Andréll J, et al. Persistence of SARS- CoV-2-specific B and T cell immunity after natural infection with severe acute respiratory responses in convalescent COVID-19 patients 6-8 months after syndrome coronavirus-2 (SARS- CoV-2) - Kentucky, USA, 2020. the infection. Med (NY) 2021;2:281e295.e4. Int J Infect Dis 2022;114:21e8. 12. Dan JM, Mateus J, Kato Y, Hastie KM, Yu ED, Faliti CE, et al. 21. Keeton R, Tincho MB, Ngomti A, Baguma R, Benede N, Suzuki A, Immunological memory to SARS-CoV-2 assessed for up to 8 et al. T cell responses to SARS-CoV-2 spike cross-recognize months after infection. Science 2021;371:eabf4063. Omicron. Nature 2022;603:488e92. 536