Development of equine antisera with high neutralizing activity against SARS-CoV-2

Dr. Gajanan Sapkal ICMR-National Institute of Virology, Pune Dr. Anil Yadav Biological E. Limited, Hyderabad. Dr. Gururaj Rao Deshpande ICMR-National Institute of Virology, Pune Dr. Pragya D. Yadav ICMR-National Institute of Virology, Pune Ms. Ketki Deshpande ICMR-National Institute of Virology, Pune Mr. Darpan Phagiwala ICMR-National Institute of Virology, Pune Dr. Rajlaxmi Jain ICMR-National Institute of Virology, Pune Dr. Anita Shete ICMR-National Institute of Virology, Pune Dr. Nivedita Gupta Indian Council of Medical Research, New Delhi Mr. Sampath Ponnuru Biological E. Limited, Hyderabad. Mrs. Kamala Palakurthi Biological E. Limited, Hyderabad. Dr. Vikram Paradkar Biological E. Limited, Hyderabad. Prof. Priya Abraham (  priya.abraham@icmr.gov.in ) ICMR-National Institute of Virology, Pune.

Although there are few drugs approved for the treatment and many under different phases of trial 3,4 , there is not yet a safe and effective vaccine available for COVID-19. Prophylactic and /or therapeutic intervention strategies would be the most appropriate and effective method for control of the rapid spread of this dreadful disease. Immunoglobulins are well-known for its therapeutic property against many diseases. These immunoglobulins are produced using the inactivated pathogen or toxins as an immunogen to immunize equines to generate hyper-immune serum containing polyclonal IgG's. The practice of administering puri ed polyclonal immunoglobulins (IgG) from hyperimmune sera of animals has been used extensively in the successful control of many viral and bacterial infections i.e. Rabies, Cytomegalovirus, Hepatitis B, Vaccinia virus, Tetanus, Botulism, Diphtheria etc. 5,6,7. A few studies have reported the equine hyperimmune sera against SARS-CoV-2 S1 RBD proteins as being potent because of generation of high titers of neutralizing antibodies 8,9 .
Passive immunization of severely ill COVID-19 patients with plasma from recovered COVID-19 patients was reported to be an effective treatment option 10,11 . The US Food and Drug Administration (FDA) issued an emergency use authorization (EUA) of convalescent plasma for the treatment of hospitalized COVID-19 cases on August 23, 2020 12 . However, obtaining the plasma from recovered patients is a di cult task and its titer and quality keep changing from one patient to other. In such situations, production of antiserum in large animals is a good alternative. Considering the need for effective therapeutics, we have developed and evaluated a SARS-CoV-2 (COVID-19) antiserum immunoglobulin (Puri ed F(ab')2 fragments) against SARS-CoV-2.

Materials And Methods
Ethical statement: The study was approved by the Institutional Animal Ethics Committee and further approved by Virus titrations were performed in Vero CCL-81 cells using tissue culture infectious dose 50% (TCID50) assay. Virus titre (TCID50/ml) was calculated by the Reed-Muench method and found to be 10 6.5 TCID50/ml.

Antigen preparation
Gamma inactivation of the virus: Gamma irradiation of the virus stock was performed using Co-60 source (24 kGy) of GC-5000 Gamma chamber (BRIT, Mumbai). This irradiated stock was again inoculated in Vero CCL-81 twice and observed for ve days to con rm the complete inactivation of the virus (Elliot et al., 1982) 14 .
Concentration of gamma-inactivated antigen: Gamma irradiated SARS-CoV-2-infected tissue culture uid was concentrated using 30 kDa lters (Pall, Germany) and further passed through 0.2 μm lters, aliquoted and stored at −80°C. Concentrated viral antigen was also aliquoted in 1 and 2 ml volumes in frosted glass bottles and further lyophilized. The lyophilized vials were stored at −20°C to be used as a source of whole virus antigen Equine Immunization: Ten 4-10 years old, healthy equines (160-200 kg in weight) that had no detectable antibodies against SARS-CoV-2, were chosen for primary immunization at Biologicals E. Ltd (Bio E). The equines were numbered HK1 to HK10, and the same numbers were used to represent the plasma obtained from individual equines accordingly. Equines were inoculated with inactivated SARS-CoV-2/VeroCCL81/P-4 antigen subcutaneously along with Freund's adjuvant. After completion of initial immunization, the equines were test bled and plasma samples were collected and anti-SARS-CoV-2 IgG was tested via ELISA and plaque reduction neutralization assay (PRNT).

Enzyme-linked immunosorbent assay (ELISA)
Speci city of antibodies raised in equines were evaluated by sandwich ELISA using inactivated SARS-CoV-2 antigen (Sapkal et al., 2020) 15 . Brie y, 96-well polystyrene microtitre ELISA plates (Nunc, Thermo Fisher Scienti c, USA) were coated with inactivated SARS-CoV-2 antigen (1:10 diluted, 100 µl/well) in 1x Phosphate-Buffered Saline (PBS) (pH 7.2), overnight at 4°C and then were blocked with a 1% BSA in 1x PBS for one hour at 37°C. The plates were washed three times with 1x PBST, pH 7.4 with 0.1 % tween-20 (PBST). To the coated plate, 100 µl of 1:100 diluted equine plasma samples were added and incubated at 37°C for one hour. After each step, the plate was washed ve times using 1x PBST. Following this, 100 µl/well of anti-horse IgG horseradish peroxidase (HRP, Sigma-Aldrich, USA) (1:16000) diluted in 1x PBST and added; and plates were incubated for one hour at 37 C. Further, 100 µl of 3, 3', 5, 5'tetramethylbenzidine (TMB, Cellbiosis) substrate was added and incubated for 10 min. The reaction was stopped by adding 100ul of 1N sulphuric acid (H2SO 4 ), and the absorbance values were measured at 450 nm using an ELISA reader (Thermo Fischer scienti c, USA). Normal horse plasma (non-immunized) was used as negative control and pooled plasma of 10 immunized animals been used as positive control. The cut-off for the assay was de ned as mean of negative control optical density plus three standard deviation (3SD).
Further to identify the binding e cacy of the antibodies, multiple dilutions of pooled plasma collected from the equines and the puri ed bulk preparations produced from the plasma were tested in ELISA. For negative control, normal horse serum (non-immunized) was added. The bound antibodies were then probed with anti-horse IgG-HRP conjugate. The end point titre was de ned as the reciprocal of the highest dilution of the sample that gives result above the cutoff 16 .
Plaque reduction neutralization test (PRNT): PRNT was performed as described by Deshpande et al., (2020) 17 . Brie y, four-fold serial dilutions of heat inactivated (56°C for 1 h) horse plasma samples were mixed with an equal amount of virus suspension containing 50-60 plaque forming units (pfu) in 0.1 ml. After incubating the mixtures at 37°C for one hour, each virus-diluted plasma sample (0.1 ml) was inoculated onto one well of a 24-well tissue culture plate containing a con uent monolayer of Vero CCL-81 cells. After incubating the plate at 37°C for one hour, overlay medium (2% CMC with 2% FBS in 2× MEM) was added to the cell monolayer, and the plate was further incubated at 37°C in 5% CO 2 for 4 days.
Plaques were observed and the plates were stained with 1% amido black for an hour. Antibody titres were de ned as the highest plasma dilution that resulted in >90 per cent (PRNT 90 ) reduction in the number of plaques.
Puri cation of SARS-CoV-2 F(ab')2 equine immunoglobulin Plasma collected from the 10 equines was pooled to produce a plasma pool and was used to produce multiple batches of puri ed F(ab')2 fragments as per established manufacturing technology at BioE at pilot scale. In the rst step, the plasma pool was diluted and the pH adjusted to 3-3.5 and enzyme pepsin was added to initiate IgG digestion. During the process, the pH and temperature was controlled to ensure complete digestion. After the completion of the enzymatic reaction, the pH of the solution was raised and heated till 55 o C to inactivate residual pepsin as well as any equine viruses potentially present in the plasma. After heat inactivation, caprylic acids were added to precipitate contaminating plasma proteins such as albumin. The precipitated proteins were removed from the solution via cloth ltration. The ltered solution containing F(ab')2 fragments was then further puri ed using ultra ltration-dia ltration and formulated into a glycine-sodium chloride buffer. The formulated bulk was ltered through a 0.2 micron lter to produce Puri ed Bulk F(ab')2 fragment immunoglobulins.
Two to three batches of puri ed bulk were pooled and mixed with the glycine-sodium chloride buffer and then sterile ltered to produce nal bulks. The nal bulk preparations were again sterile ltered in an online manner on an automated vial lling line and lled into 2R glass vials stopper with rubber bungs and sealed with ip-off aluminium caps to produce nal lot product suitable for clinical evaluation. Nonreducing SDS-PAGE gels, using the buffer system described by Laemmli (1970) 18 , were used to monitor the digestion process.
Overall scheme of immunization strategy and equine hyperimmune globlulin production has been represented in Figure-

SARS-CoV-2 speci c IgG and neutralizing antibody response
Ten healthy horses were immunized with inactivated SARS-CoV-2 virus subcutaneously and after 21 days of immunization, plasma samples were tested by anti SARS-CoV-2 IgG ELISA and PRNT for the detection of nAb titres. The results of the plasma samples obtained from HK1 to HK10 indicated the presence of SARS-Co-V2 speci c IgG antibodies as detected in ELISA with neutralizing capacity. (Figure 2).
Further, the neutralization activity of the plasma pooled from the ten equines after four rounds of immunization were tested for the presence of nAb by PRNT. The nAb titre of the pooled plasma was >4096 (last dilution of antibody tested). From the pooled plasma, seven lots of puri ed bulks were prepared and the puri ed bulk batches were tested for anti-SARS-Co-V2 binding IgG antibody by ELISA and nAb by PRNT to recognize the variation of antibody titres between the batches, the results are represented in Figure-3.

Puri cation and characterization of F(ab')2 from pooled plasma:
Two or three of the puri ed bulk preparations were pooled and then sterile ltered to generate three nal bulk preparations. Each nal bulk was further sterile ltered and lled into 2R glass vials on an automated lling line. During the lling process, vials were collected at the beginning, middle and end stages. Each nal bulk and the lled vials produced from the nal bulk were tested for PRNT90 titers as summarized in Figure 4.
The nAb titers are consistent from Final Bulk to Final Lot ( lled vials) across entire lling operation during all three batches.
Preparation of F(ab')2 : Purity of F(ab')2 of Final bulk and Final lot samples was assessed by SDS-PAGE to check key plasma impurities such as albumin. The SDS-PAGE pro le shows the F(ab')2 fragment band at molecular weight of approximately 100 kDa. Albumin was added in multiple lanes at 0.5 to 8 mcg quantity corresponding to 0.5% to 8% of relative impurity content in the nal bulk and nal lot samples ( Figure 5).

Measurement of IgG titres in puri ed plasma by Anti-SARS Co-V2 ELISA
The results of ELISA indicated the good speci city of the antibodies raised in equines against SARS-CoV-2. The binding e cacy of the antibodies was also evaluated by this ELISA which showed a capacity of recognizing antigen at the highest dilution of 1:81920 of equine sera ( Figure 6).

Quality control assessment of the puri ed products
The quality of three batches of the nal bulk and nal lot were checked as per the standard criteria. The samples were checked for appearance, pH, total protein content, osmolality, purity, molecular size distribution, albumin percentage and a few others. The pH of the nal bulk and lot for all the three batches were between 6.37 to 6.49 for bulk and 6.45 to 6.55 for the nal lot. The overall purity of the F(ab')2 fragments produced from plasma processing was evaluated by SDS-PAGE and SEC-HPLC (data not shown) and all the three batches of nal bulk and the nal lots produced from these bulks showed consistent purity of 99%+ and minimal content of impurities such as aggregates and other plasma proteins such as albumin. The detailed descriptions of the quality check criteria are in the appendix (Supplementary tables 1 and 2). Pooled plasma and the three nal lot batches produced from the plasma were tested by a validated PCR method for a panel of equine viruses and all were absent indicating appropriate inactivation during the process (data not shown).

Discussion
In the current state of the pandemic, due to the unavailability of approved speci c vaccines and drugs for treatment of SARS-Co-V2, an urgent need of therapeutic strategies are required. Human convalescent plasma has not met with the desired results 8 . Puri ed immunoglobulins obtained from hyper-immune equine sera has been an effective and time-tested approach in various infections such as diphtheria, tetanus, rabies and bites from snakes, scorpions, arachnids and more recently SARS-CoV-1, MERS-CoV, Ebola and avian in uenza virus 23,24,25,26 .
The enzyme puri ed equine F(ab')2 without FC region further reduces side effects and makes it more suitable for use as a therapeutic agent to neutralize the pathogen . However, being a heterologous protein it may be susceptible to immune resistance by the host; hence a large therapeutic dose may be required for human administration. Further, the equine antibodies acts as broad-spectrum antiviral drugs due to its property of multi target action 27. In this study, we report the preparation of equine hyper-immune sera to demonstrate their protective e cacy against SARS-CoV-2 virus using an in vitro live virus neutralization assay.
The antiserum was prepared by injecting inactivated whole virus antigen in horses subcutaneously for a period of 21 days. The resulting nAb titres in the plasma of the immunized equines displayed high titres against SARS-CoV-2. Since heterologous antisera can cause adverse effects as reported earlier in case of SARS-CoV-1, these antibodies were fragmented by pepsin to purify the F(ab')2 fragments. F(ab')2 from three batches of hyper-immune sera generated from ten equines showed similar protective effects against live virus. As was evident during the study, the neutralizing activity of the antibodies was preserved during the enzymatic digestion process that yielded highly puri ed and effective F(ab')2 fragments. The purity of the antisera was recorded as 99 % and the neutralizing titres above 20,000.
Our results are in agreement with the other studies on equine antisera for SARS-CoV-2 which reported generation of high nAb in horses against receptor binding domain of the spike protein of the virus. This study suggests promising e cacy and therapeutic potential of equine hyper immune sera against SARS-CoV-2. Equine hyper-immune serum overcomes the challenge of limited availability of convalescent plasma from recovered patients. Monoclonal antibodies on the other hand are laborious and expensive to generate. Equine antiserum is now known to be safer, since it is devoid of the Fc domain of antibody and can be prepared in bulk at a lower cost.

Conclusion
The study provides evidence of the potential of generating highly puri ed F(ab')2 from equines against SARS-CoV-2 that can demonstrate consistent and high neutralization activity. Further, in-vivo testing for e cacy of this indigenously developed, cost effective product will pave the way to clinical evaluation.
Additionally, being a donor independent method, this may prove as an e cient alternative to convalescent plasma for treatment of COVID19 patients. Figure 1 Scheme of immunization strategy and equine hyperimmune globlulin production