Introduction

In order to design and construct the neuraminidase-boosted influenza virus, we utilized engineering principles. This engineering design cycle comprises four stages: design, construction, testing, and learning.(Figure 1).

Figure 1. Engineering schematic diagram. A cycle includes design, build, test, and learning.The image used in this content has been sourced from https://2022.igem.wiki/peking/engineering.

Rescued the virus whose neuraminidase was extended.

Overview: In our project, we aim to construct a neuraminidase-boosted influenza virus. We have employed two methods to modify the influenza virus. One of these methods involves inserting amino acid fragments from other influenza viruses into the stalk region of neuraminidase, thereby extending neuraminidase and making it more prominent on the surface of the influenza virus. Throughout this process, we have undergone three rounds of engineering rounds.

First round

1.Design: Based on the A/PR/8 (H1N1) influenza virus strain, Broecker et al. inserted amino acid fragments from the A/California/04/2009 (Cal09) virus and the A/New York/61/2012 (H3N2) virus into the stalk region of the PR8 neuraminidase and constructed Ins15 virus with 15 amino acids inserted and Ins30 virus with 30 amino acids inserted (Figure 1.1). This successful modification induced a better immune response against the neuraminidase in mice ^[1]. Since the influenza virus we utilized was the A/WSN/1933 (H1N1) strain, which shares a strong affinity with the PR8 influenza virus, we intended to insert the same sequence into the neuraminidase of the WSN strain. Referring to the comparison of neuraminidase protein sequences between the WSN strain and the PR8 strain by Hiti AL et al. (Fig. 1.2) ^[2], we also inserted the same 15 amino acids and 30 amino acids at similar sites in the WSN strain.

Figure 1.1 Influenza virus designed to extend the neuraminidase stack domain ^[1]

Figure 1.2 Sequence comparison of neuraminidase between PR8 strain and WSN strain ^[2].The position marked by the red line in the figure is the insertion point

2.Construction: The virus rescue system we utilized was the 12-plasmid system provided by Academician Gao Fu. The A1-A8 plasmids were employed for the transcription of viral RNA, while the A9-A12 plasmids were used to express the PA, PB1, PB2, and NP proteins to simulate the state of virus-infected cells. Within the virus rescue system, the A6 plasmid was utilized for the transcription of RNA corresponding to the viral neuraminidase. We aimed to insert the 15 amino acids and 30 amino acids, as shown in Figure 1, at the 61st position of the neuraminidase. Consequently, we constructed the A6-Ins15 plasmid and the A6-Ins30 plasmid through homologous recombination, based on the A6 plasmid. Sequencing results confirmed the successful insertion of the target fragment (Figure 1.3).

Fig. 1.3 Sequencing Results of A6-Is15 Plasmid and A6-Is30 Plasmid

3.Test: Once the mutant plasmid was successfully constructed, we proceeded to rescue the virus using the mutant plasmid. HEK293T cells were transfected with the mutant plasmid using PEI transfection reagent. When approximately 90% of the cells had died, we collected the supernatant containing the virus and removed cell debris through centrifugation. Subsequently, we introduced a small amount of the virus solution into a fresh batch of HEK293T cells and harvested the total RNA after 4 hours for reverse transcription to obtain cDNA. The cDNA was then amplified through PCR to target the NA protein. The amplified cDNA was sent to a sequencing company for mutation detection. The sequencing results revealed the successful rescue of the wild-type influenza virus and the Ins15 mutant virus, but not the Ins30 mutant virus (Figure 1.4). However, the Ins15 mutant virus did not effectively induce cytopathic effects (CPE) in cells (Fig. 1.5). In other words, directly introducing the same mutations from the PR8 strain into the WSN strain to extend the neuraminidase is not a feasible approach.

Figure 1.4 Results of viral neuraminidase RNA sequencing

Figure 1.5 CPE Effect Results of Wild Influenza Virus, Ins15 Influenza Virus and Ins30 Influenza Virus

4.Learning: Upon reviewing the entire design, we discovered significant differences in the sequence of the stalk region between the WSN and PR8 strains, particularly at our insertion position. This difference may be the reason for our failure. Consequently, we conducted further literature research and came across studies on the modification of the neuraminidase stalk region in the WSN strain. In their investigation of the biological activity of the influenza virus neuraminidase stalk length, Castrucci MR et al. constructed influenza viruses with varying neuraminidase stalk lengths based on the WSN strain. These modifications included deleting 24 amino acids (SD0), deleting 9 amino acids (SD9), inserting 14 amino acids (SA14), and inserting 28 amino acids (SA28) (Figure 1.6). The researchers observed that the stalk length had an impact on virus replication efficiency: the longer the stalk, the better the replication effect^[3].

Figure 1.5 CPE Effect Results of Wild Influenza Virus, Ins15 Influenza Virus and Ins30 Influenza Virus

Figure 1.6 Design of neuraminidase stack regions with different lengths ^[3]

Second round

1.Design:Drawing from the research conducted by Castrucci MR et al., we attempted to rescue four mutant viruses they had constructed in HEK293T cells.

2.Construction: Similar to the initial construction process, our focus remained on modifying the A6 plasmid within the twelve-plasmid system. We performed deletions and insertions to generate the following mutants: A6-SD24, which involved the deletion of amino acids 38-61; A6-SD9, which involved the deletion of amino acids 43-51; A6-SI14, which entailed the insertion of 14 amino acids from A/Tokyo/67 (H2N2) at position 42; and A6-SI28, which involved the insertion of 28 amino acids from A/Tokyo/67 (H2N2) and A/Tern/Australia/G70C/75 (H11N9) at position 42. Through point mutation or homologous recombination, we successfully constructed these plasmids. Sequencing results confirmed the successful deletion or insertion of the target fragments (Figure 1.7).

Figure 1.7 Comparison Diagram of Mutant Plasmid Sequencing Results

3.Test: Once the mutant plasmids were successfully constructed, we attempted to rescue the viruses using these mutants. HEK293T cells were transfected with the mutant plasmids using PEI transfection reagent. After observing cell death in approximately 90% of the cells, we collected the supernatant containing the viruses and removed cell debris through centrifugation. Subsequently, we added a small amount of the virus solution to a fresh batch of HEK293T cells and extracted total RNA 4 hours later for reverse transcription to obtain cDNA. The cDNA was then amplified through PCR to target the NA protein, and the resulting PCR products were sent to a sequencing company to detect any mutation sequences. The sequencing results confirmed the successful rescue of the wild-type influenza virus and the four mutant viruses. Additionally, the mutant viruses effectively induced cytopathic effects (CPE) in cells (Fig. 1.8 and 1.9). Encouraged by these results, we attempted to generate a larger quantity of mutant viruses and performed a hemagglutination inhibition assay using chicken red blood cells to obtain a rough quantification of the viruses. We observed significant inhibitory effects on chicken red blood cell agglutination for both the SD9 and SI14 mutant viruses, while SD24 and SI28 did not exhibit such effects (Figure 1.9). Furthermore, upon absolute quantification of the virus RNA, we discovered that the virus titers of SD24 and SI28 were much lower than those of the wild-type, SD9, and SI14 viruses (Figure 1.10). According to the literature, we were aware that SD24 was a highly defective virus, and its low viral titer was expected. However, SI28, which was reported as the most efficient virus in the literature, displayed a low virus titer, indicating potential issues with our design.

In this round, we successfully generated four mutant viruses: SD24, SD9, SI14, and SI28, although SI28 did not meet our expectations.

Figure 1.8 CPE Effect Results of Wild Influenza Virus, SD9 and SI14 Influenza Virus

Figure 1.9 CPE effect of SI28 and SD24 mutant influenza viruses

Figure 1.10 Results of chicken red blood cell hemagglutination test.

The figure below is the opposite of the figure above, SI28, SI14, SD9, SD24 and WT from top to bottom.

Figure 1.11 Results of virus titer

4.Learning: Upon rechecking the sequence, we identified the problem. During the design of the SI28 mutant plasmid, we inadvertently copied the base sequence of 14 amino acids from A/Tern/Australia/G70C/75 (H11N9) incorrectly. Instead of copying amino acids at positions 58-71, we mistakenly copied amino acids at positions 52-65. As there was no coding shift mutation, we initially failed to detect this error.

Third round

1.Design: Considering the previous realization, we proceeded to redesign A6-SI28-new to rectify the errors identified.

2.Construction: We obtained the corresponding primers and performed a re-synthesis. By employing homologous recombination, we successfully constructed the aforementioned plasmid. Sequencing results confirmed the successful insertion of the target fragment.

Figure 1.12 Comparison Diagram of SI28 new Sequencing Results

3.Test: After successfully constructing the mutant plasmid, we proceeded to rescue the virus using this mutant construct. HEK293T cells were transfected with the mutant plasmid using PEI transfection reagent. Following the observation of cell death in approximately 90% of the cells, we collected the virus-containing supernatant and removed cell debris through centrifugation. Subsequently, we added a small amount of the virus solution to a fresh batch of HEK293T cells. After 4 hours, we extracted total RNA and performed reverse transcription to obtain cDNA. We then amplified the cDNA corresponding to the NA protein through PCR and sent it to a sequencing company to detect any mutation sequences. The sequencing results confirmed the successful rescue of the SI28-new mutant virus. However, due to the virus packaging round (requiring at least one week), we were unable to complete the large-scale preparation of the SI28-new mutant virus before the submission of the wiki. Therefore, we were unable to confirm the successful rescue of the mutant virus at the protein level. We plan to continue this experiment before the jamboree to address this limitation.

4.Learning: During the plasmid construction process, it is crucial to exercise caution to avoid mistakenly inserting fewer or additional fragments. Multiple checks and confirmations should be conducted before proceeding with the experiment to prevent unnecessary time wastage. This mistake also highlighted the highly variable nature of the neuraminidase stalk region.

Rescued the virus whose packaging signals of hemagglutinin and neuraminidase were swapped.

Overview: Our project focuses on constructing a neuraminidase-boosted influenza virus. We have employed two methods to modify the influenza virus, one of which involves reversing the expression levels of neuraminidase and hemagglutinin by exchanging the packaging signals of their vRNA segments. This approach aims to enhance the display of neuraminidase on the surface of the influenza virus. Throughout the project, we have undergone three rounds of engineering rounds.

First round

1.Design: Allen Zheng et al., based on A/PR/8 (H1N1) influenza virus strain, rescued the influenza virus expressing more neuraminidase and less hemagglutinin by exchanging the packaging signals at the 5 'and 3' ends of the hemagglutinin and neuraminidase vRNA segments. The vaccine prepared by this influenza virus triggered a stronger immune response against neuraminidase in mice (Figure 2.1)^[4].

Figure 2.1 Exchange HA and NA packaging signal design ^[4]

Drawing on the ideas of this article, we also hoped to build a WSN mutant virus that exchanges HA and NA packaging signals. To accomplish this, we conducted an extensive literature review to identify the specific sequences of the HA and NA packaging signals^[5][6] (Figure 2.2, Figure 2.3).

Figure 2.2 HA Packaging Signal ^[5].The part marked with yellow in the figure indicates the length of packaging signal 5 'end packaging signal: 33nt non coding area+9nt coding area; 3 'packaging signal: 45nt non coding area+80nt coding area

Figure 2.2 NA Packaging Signal ^[6].The part marked with yellow in the figure indicates the length of packaging signal 5 'end packaging signal: 19nt non coding area+183nt coding area; 3 'packaging signal: 28nt non coding area+157nt coding area

2.Construction:In the twelve-plasmid system, HA corresponds to the A4 plasmid, and NA corresponds to the A6 plasmid. Based on the aforementioned research, we designed multiple pairs of primers to obtain the packaging signal from the A4 and A6 plasmids. Using the homologous recombination method, we added the packaging signal from each other on both sides of the coding region of HA and NA. This allowed us to construct the A4-NAps plasmid and A6-HAps plasmid, and we also mutated the ATG in the coding region of the packaging signal to TTG. The sequencing results confirmed the successful insertion of the target fragment.

Figure 2.3 Schematic Diagram of A4 NAPs and A6 HAps

3.Test:Following the successful construction of the mutant plasmid, we attempted to rescue the virus using this modified construct. HEK293T cells were transfected with the mutant plasmid using PEI transfection reagent. Upon observing approximately 90% cell death, we collected the virus-containing supernatant and removed cell debris through centrifugation. Subsequently, we added a small amount of the virus solution to a fresh batch of HEK293T cells. After 4 hours, we extracted total RNA for reverse transcription to obtain cDNA. We then performed PCR amplification using vRNA primers specific to the NA and HA proteins. The amplified cDNA samples were sent to a sequencing company to detect any mutation sequences. Unfortunately, the sequencing results from the samples we submitted all indicated wild-type sequences.Furthermore, due to the shorter vRNA sequence of NA (1409bp) compared to HA (1775bp), the results of cDNA amplification using the corresponding vRNA primers also revealed the lack of successful mutations (Figure 2.4). Moreover, after exposing HEK293T cells to the virus supernatant, no noticeable cytopathic effects (CPE) were observed (Figure 2.5).

Figure 2.4 Agarose gel electrophoresis of exchange packaging signal virus vRNA PCR products

Figure 2.5 CPE effect diagram of exchange packaging signal virus

4.Learning:Further consideration of the literature led us to recognize that the packaging signal can be divided into two components: the non-coding region and the partial coding region. After exchanging the packaging signals, we neglected to remove the packaging signals in the coding regions of the HA and NA proteins. This oversight may have had a negative impact on the successful rescue of the virus.

Second round

1.Design: Based on the first round of analysis, we used the codon optimization website (https://www.vectorbuilder.cn/tool/codon-optimization.html)1. to introduce synonymous mutations in the packaging signals of the coding regions of the HA protein and NA protein.

2.Construction: Using the results of the synonymous mutation analysis, we designed a series of primers and modified the A4-NAps plasmid and A6-HAps plasmid through homologous recombination. This allowed us to successfully introduce synonymous mutations in the packaging signal of the coding region, resulting in the construction of the A4-NAps(mut) plasmid and A6-HAps(mut) plasmid. The sequencing results confirmed the successful introduction of synonymous mutations.

3.Testing: After successfully constructing the mutant plasmids, we attempted to rescue the virus using these plasmids. HEK293T cells were transfected with the mutant plasmids using PEI transfection reagent. When approximately 90% of the cells died, we collected the virus-containing supernatant and removed cell debris through centrifugation. Subsequently, a small amount of the virus solution was added to another batch of HEK293T cells, and the total RNA was extracted after 4 hours for reverse transcription to obtain cDNA. PCR was performed to amplify the vRNA of the NA and HA proteins, and the amplified products were sent to a sequencing company for mutation analysis. The sequencing results revealed that we were unable to successfully rescue the mutant virus with exchanged packaging signals. Instead, the wild-type influenza virus was still present.

Figure 2.6 Agarose gel electrophoresis of PS(mut) virus vRNA PCR products

4.Learning: We reexamined the A4-NAps(mut) and A6-HAps(mut) plasmids and identified the cause. In the A4-NAps(mut) plasmid, we had mutated the two initiation codons ATG in the 5'-end packaging signal. However, upon further review, we discovered two additional frameshift initiation codons, ATG. These two ATGs may be the cause of the failure.

Third round

1.Design: We made point mutations in the two initiation codons mentioned above.

2.Construction: Two pairs of point mutation primers were designed, and the remaining two ATGs were successfully mutated to TTG in A4-NAps(mut). The sequencing results confirmed the successful introduction of the point mutation.

3.Testing: Due to a lack of time, we did not test the new plasmids before the wiki submission. We will continue this part of the experiment before the jamboree.

4.Learning: There is always a reason why experiments don't meet expectations, and by carefully examining them, we can discover interesting aspects.

Exploration of Western blotting conditions

Overview: Western blotting can separate proteins with different molecular weights and semi-quantify them. We hoped to characterize the extended neuraminidase by this method. In the process of exploring suitable conditions, we adopted the principle of engineering. In this process, we went through two rounds.

First round

1.Design: When the virus was prepared in large quantities and concentrated, we extracted the total RNA by attacking the virus and obtained the cDNA by reverse transcription. After PCR, the mutation was confirmed at the gene level, and it was also necessary to confirm the mutation at the protein level. Through literature research, we used Western blotting to characterize protein.

2.Construction: We tried using the methods and parameters commonly used in the laboratory. For the characterization of neuraminidase with different lengths, because it does not need accurate quantification, we directly use the stock solutions of different viruses to prepare samples ,which need to be diluted to the same order of magnitude according to the results of chicken red blood cell hemagglutination experiment. After boiling the sample at 100℃ for 10 min, 12% preformed glue was used for electrophoresis at 140V for 60min. After electrophoresis, transfer membrane at 100V for 60min. Incubate the primary antibody and secondary antibody, and then expose it.

3.Test: After exposure, we did not distinguish neuraminidase with different lengths obviously (Figure 3.1).

Fig. 3.1 Western blotting chromogram of neuraminidase with different lengths.

4.Learning: The average molecular weight of amino acids is calculated, 1kD is equivalent to about 8 amino acids. If the wild-type neuraminidase is taken as a reference, the neuraminidase of SD24 is about 3kd smaller, that of SD9 is about 1kd smaller, that of SI14 is 1.7kd larger and that of SI28 is 3.5kD larger. Among them, SD9 and SI14 are not easy to distinguish from wild type. Western blotting can be roughly divided into the following steps: sample preparation, SDS-PAGE electrophoresis, membrane transfer, antibody incubation and exposure. Among them, the step of determining whether protein with different lengths can be distinguished is mainly SDS-PAGE electrophoresis. So, we tried to explore the different conditions of SDS-PAGE electrophoresis.

Second round

1.Design: Fortunately, according to the experiment in the first round, we determined the approximate position of neuraminidase (between 55 KD and 70 KD), so we can adjust the electrophoresis conditions by the distance between marker bands near neuraminidase. We designed different SDS-PAGE electrophoresis conditions, and the main transformation parameters were voltage and time.

2.Construction: We tried different electrophoresis conditions with 12% preformed glue.

3.Testing: After electrophoresis, membrane transfer, antibody incubation and exposure were carried out under uniform conditions. The following are the results of each condition. You can see that the bands gradually separate with the exploration of conditions.

Fig. 3.2 Western blotting chromogram under different electrophoresis conditions

4.Learning: 4.Adjusting the experimental parameters according to the experimental purpose is a necessary ability in scientific research, and it is forbidden to mechanically copy the experiment.

Reference

[1]Broecker F, Zheng A, Suntronwong N, et al. Extending the Stalk Enhances Immunogenicity of the Influenza Virus Neuraminidase. J Virol. 2019;93(18):e00840-19. Published 2019 Aug 28. doi:10.1128/JVI.00840-19

[2]Hiti AL, Nayak DP. Complete nucleotide sequence of the neuraminidase gene of human influenza virus A/WSN/33. J Virol. 1982;41(2):730-734. doi:10.1128/JVI.41.2.730-734.1982

[3]Castrucci MR, Kawaoka Y. Biologic importance of neuraminidase stalk length in influenza A virus. J Virol. 1993;67(2):759-764. doi:10.1128/JVI.67.2.759-764.1993

[4]Zheng A, Sun W, Xiong X, et al. Enhancing Neuraminidase Immunogenicity of Influenza A Viruses by Rewiring RNA Packaging Signals. J Virol. 2020;94(16):e00742-20. Published 2020 Jul 30. doi:10.1128/JVI.00742-20

[5]Watanabe T, Watanabe S, Noda T, Fujii Y, Kawaoka Y. Exploitation of nucleic acid packaging signals to generate a novel influenza virus-based vector stably expressing two foreign genes. J Virol. 2003;77(19):10575-10583. doi:10.1128/jvi.77.19.10575-10583.2003

[6]Fujii Y, Goto H, Watanabe T, Yoshida T, Kawaoka Y. Selective incorporation of influenza virus RNA segments into virions. Proc Natl Acad Sci U S A. 2003;100(4):2002-2007. doi:10.1073/pnas.0437772100