Serum from COVID-19 patients early in the pandemic shows limited evidence of cross-neutralization against variants of concern

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) results in a variety of clinical symptoms ranging from no or mild to severe disease. Currently, there are multiple postulated mechanisms that may push a moderate to severe disease into a critical state. Human serum contains abundant evidence of the immune status following infection. Cytokines, chemokines, and antibodies can be assayed to determine the extent to which a patient responded to a pathogen. We examined serum and plasma from a cohort of patients infected with SARS-CoV-2 early in the pandemic and compared them to negative-control sera. Cytokine and chemokine concentrations varied depending on the severity of infection, and antibody responses were significantly increased in severe cases compared to mild to moderate infections. Neutralization data revealed that patients with high titers against an early 2020 isolate had detectable but limited neutralizing antibodies against newly circulating SARS-CoV-2 variants of concern. This study highlights the potential of re-infection for recovered COVID-19 patients.


37
In December 2019, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was first reported in 38 the city of Wuhan, Hubei province, China, causing variably severe respiratory tract pathology termed 39 coronavirus disease 2019 . COVID-19 is often a mild disease associated with low-grade fever 40 and loss of taste and smell. However, critical cases of COVID-19 do occur, and are characterized by 41 severe pneumonia and acute respiratory distress syndrome (1) leading to organ failure and death (2). As 42 of October 20 th 2021, over 241 million cases have been reported worldwide, and over 4.9 million people 43 have died of COVID-19 (https://coronavirus.jhu.edu/map.html).

44
The spectrum of disease caused by SARS-CoV-2 ranges from no or mild to critical. Mild to 45 moderate cases are characterized by mild symptoms ranging to mild pneumonia and account for up to 46 81% of infections. Severe cases account for 14% of cases, which involve dyspnea, hypoxia, or greater 47 than 50% lung involvement as determined by imaging. Five percent of patients are deemed critical 48 based on conditions of respiratory failure, shock, or multiorgan system dysfunction (3,4). In many 49 severely affected patients, SARS-CoV-2 infection triggers an overactive immune response known as a 50 "cytokine storm." Immune cells produce high levels of inflammatory cytokines leading to systemic shock 51 and death (5). As such, cytokines have been studied extensively in the context of SARS-CoV-2 infection 52 and have been found to be central to the pathophysiology of COVID-19 (6, 7).

53
A thorough understanding of appropriate immune responses is vital to the development of 54 effective medical intervention strategies and vaccines. Besides cytokine and chemokine production 55 following infection, antibodies generated by COVID-19 patients have been studied and reported in 56 detail. Infection with SARS-CoV-2 has been found to induce non-class-switched, class-switched, and 57 neutralizing antibodies in immunocompetent patients (8)(9)(10)(11)(12). The long term stability of the antigen-58 specific and neutralizing antibody response has been found to be up to 13 months in patients (13-16).

59
Pre-existing antibody populations may also contribute to disease severity such as autoantibodies to type 60 I interferons (17). As SARS-CoV-2 mutates, changes to the sensitivity of pre-exisitng neutralizing 61 antibody populations may be effected (18). As such, the beta and delta variants both have displayed 62 decreased sensitivity to pre-existing neutralizing antibodies (15,(19)(20)(21).

63
In this study, we evaluated 131 serum and plasma samples from 55 COVID-19 patients alongside 64 serum and plasma from 20 uninfected patients for the presence of 38 cytokines and chemokines, anti-4 severity. We also found that COVID-19 patients exhibit increased titers of antigen-specific IgG and 68 neutralizing antibody titers compared to uninfected individuals. Furthermore, we determined that the 69 neutralizing activity of our sample cohort extended to three new SARS-CoV-2 variants of concern (VOC), (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 5 27 (49%) had more than one specimen assessed. Based on review of clinical charts and according to CDC 99 criteria (www.cdc.gov/coronavirus/2019-ncov), patients were grouped into three illness severity 100 categories: Mild to moderate (mild symptoms to mild pneumonia), severe (dyspnea, hypoxia or more 101 than 50% lung involvement on imaging), and critical (respiratory failure, shock or multiorgan system 102 dysfunction). Of the normal samples, volunteers who provided a one-time sample were an average age 103 of 44 years (range 26-89 years) and 15 (75%) were female. All samples, including patients and 104 volunteers, were deidentified and assigned study-specific identifiers to protect patient confidentiality.

105
Samples were then aliquoted and frozen at -80°C until shipment to the Rocky Mountain Laboratories for 106 analysis. Samples were g-irradiated (4 MRad) to inactivate potential infectious pathogens upon receipt

129
Neutralizing antibody assay 130 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 and plasma samples were heat-inactivated for 30 minutes at 56°C. They were diluted 1:10 and then 1:2 132 for subsequent dilutions. SARS-CoV-2 virus stocks were diluted to 2,000 TCID50/ml and 70 µl was then 133 added to each well of diluted sample. Following a one-hour incubation at 37°C, the serum-virus mixture 134 was transferred to 96-well plates containing high-passage Vero E6 cells. After six days, cytopathic effect 135 (CPE) was read. The virus neutralization titer was determined to be the lowest concentration of serum 136 antibody where CPE was not observed.

138
Statistical analysis 139 Statistical analysis was performed using Prism 8. Statistically significant differences between groups for 140 cytokines were determined using one-way ANOVA; IgG and neutralzing titers were evaluated applying

141
Mann-Whitney test. Significance is indicated as follows: p<0.0001 (****), p<0.001 (***), p<0.01 (**) and We received 111 patient serum and plasma samples that were categorized according to CDC guidelines 148 into mild to moderate and critical cases. In addition, we obtained 20 serum and plasma samples from 149 healthy adult volunteers designated "normal" controls in our studies. We first sought to determine the

161
105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021.  had a mild to moderate infection produced RBD-specific IgG, and the data was statistically significant 181 compared to healthy controls ( Figure 3A). In contrast, critically-infected patients developed high titers of 182 RBD-specific IgG, which was significantly greater than both healthy controls and those with mild to 183 moderate COVID-19. The presence of RBD-specific IgG did not predict disease outcome, as there was no 184 significant difference in titers between patients that recovered from infection and those that did not 185 ( Figure 3B).

186
Three of the healthy control serum and plasma samples contained SARS-CoV-2-specific IgG 187 without COVID-19 medical history. We postulated that the presence of IgG may not translate to the 188 ability to neutralize SARS-CoV-2. Therefore, we assessed the serum and plasma for neutralizing 189 antibodies against SARS-CoV-2. The majority of control serum and plasma did not contain detectable 190 levels of neutralizing antibodies, with the exception of one patient who exhibited a detectable, albeit 191 very low, titer ( Figure 4A). However, mild to moderate infection led to the production of significantly 192 higher levels of neutralizing antibody titers compared to controls. Patients that were critically infected 193 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ; https://doi.org/10.1101/2021.11.10.468174 doi: bioRxiv preprint 8 exhibited high titers of neutralizing antibodies, approximately two logs greater than healthy controls and 194 one-and-a-half logs higher than patients with mild to moderate disease.

195
Some COVID-19 patients that participated in this study were sampled repeatedly over the 196 course of infection and recovery. Therefore, we determined whether their neutralizing antibody titers 197 remained stable over time. We found that these patients continued to produce neutralizing antibodies

205
Therefore, we tested serum and plasma samples that contained high titers of neutralizing antibodies 206 against the original virus, but this time we assessed neutralizing antibodies against the 3 VOC. We found 207 that the serum and plasma contained antibodies capable of neutralizing these SARS-CoV-2 variants, but 208 the antibody titers were significantly lower than the titers against the original virus ( Figure 4B).

209
Interestingly, the titers against the Alpha variant were significantly higher than those for the Beta and

218
Understanding the difference between an immune response that leads to recovery from infection and 219 one that leads to a negative outcome (aberrant) is essential in the design of treatments and vaccines, 220 which are necessary to bring an end to this devastating pandemic.

221
One avenue taken by investigators has been to explore the presence or absence of various 222 cytokines and chemokines in patient serum. Severe disease has been associated with an aberrant 223 immune response termed "cytokine storm," which is characterized by an overactivation of the immune 224 9 system leading to exaggerated levels of cytokines released into the circulation. Multiorgan dysfunction 225 and failure associated with septic shock can be fatal (5, 25). In contrast, those with mild disease exhibit 226 functional immune responses characterized by appropriate levels and types of cytokines, leading to 227 disease resolution (7). One cytokine that has been highlighted amongst research studies is IL-6. For

232
Interestingly, our data show that moderate levels of key cytokines and chemokines are evident 233 in mild to moderate cases of COVID-19. It is the "Goldilocks" phenomenon: too much or too little of 234 some cytokines is not good; rather, the levels must be "just right". In support of this concept, Yang et al.

235
observed that serum IL-1β, IL-1Ra, IL-6, IL-7, IL-10, IP-10, and TNF-α are all important in classifying 236 COVID-19 cases into mild, moderate, and severe (30). This study also found that IP-10 was significantly 237 higher in severe cases of COVID-19 compared to mild cases. Another study found that IP-10 levels were 238 highest in patients that required ICU admission (31). Our study supports the previous work finding very 239 high levels of IP-10 in the serum of patients who succumbed to SARS-CoV-2 disease.

248
Although presence of IgG is evidence of an effective immune response, it is important to 249 decipher whether these antibodies are capable of neutralizing virus. It has been found that not all 250 recovered COVID-19 patients develop sufficient neutralizing antibody titers (12). Our study showed that 251 while SARS-CoV-2 cross-reactive IgG antibodies were present in control patients, these antibodies were 252 not capable of neutralizing SARS-CoV-2. A recent study suggests that exposure to a seasonal coronavirus 253 can induce the production of antibodies against SARS-CoV-2, but these antibodies are not protective 254 against the virus (34). Therefore, detection of IgG alone cannot always predict protection from 255 reinfection. This is an important distinction that clinicians need to make when examining data from 256 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ; https://doi.org/10.1101/2021.11.10.468174 doi: bioRxiv preprint recovered patients. Optimally, a test to determine the presence of specific neutralizing antibodies 257 would be more informative than our current ELISA, which only detects antibodies that are specific for 258 SARS-CoV-2.

259
A key feature of SARS-CoV-2 is its high mutation rate (35,36). Selective pressures acting on the 260 virus have led to mutations that allow the virus to spread more efficiently and to evade host immune 261 responses (37). Fortunately, we and others have found that there appears to be some albeit limited

265
It is important to note that the cross-reactive neutralizing potential from the original SARS-CoV-2 was 266 significantly less for all three of the VOC tested. However, recent studies provide hope that even low 267 levels of neutralizing antibodies will lead to better outcomes after re-infection with VOC for patients

279
Our study provides additional support for the growing body of literature examining human 280 COVID-19 serum samples. Our data supports established work that increased levels of IL-6 and IP-10 281 contribute to enhanced disease phenotype. In addition, our study highlights the importance of both the 282 antigen-specific antibody response and its functionality to neutralize emerging VOC. The more fully we 283 understand effective immune responses to this pathogen, the greater our ability to successfully treat 284 those who are infected, and vaccinate those we hope to protect against infection.

305
All data is available in the manuscript. 306 307 308 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ; https://doi.org/10.1101/2021.11.10.468174 doi: bioRxiv preprint  105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021   105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021   (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021.  (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC  105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ; https://doi.org/10.1101/2021.11.10.468174 doi: bioRxiv preprint 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ; https://doi.org/10.1101/2021.11.10.468174 doi: bioRxiv preprint 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted November 12, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021