Adaptation in polymerase basic protein 1 family from influenza A virus to climate change ()
1. INTRODUCTION
Although climate change imposes serious problems on humans and other species [1], more and more attention is now being paid to the adaptation of different species to climate change at different levels [2-4]. Actually, this type of adaptation goes through mutations, because mutations are the base for either functional or structural adaptation. Nevertheless, adaptation makes a species more suitable to climate change and makes a species survive climate change.
Although mutations benefit a species to adapt to climate change, these mutations could be harmful to other species. This is plausible because mutations would not limit their effects only to adaptation to climate change. Accordingly, any mutation in virus could lead to unexpected effects on human health, for example, unpredictable mutations in influenza A viruses constantly threaten humans with possible deadly pandemics/epidemics. Indeed, not only the mutations whose aim is to adapt to climate change but also the mutations whose aim is to adapt to changed ecosystems, where host-pathogen interactions may change as well [5,6], should be taken into consideration.
On the other hand, virus adaptation is different from the adaptations of many other species because virus lives in host cells with relatively favorable and stable environments. With respect to influenza A virus, climate changes alter the normal pattern of bird migrations, which interrupts the transmission cycle of avian influenza A virus [7], and also influenza A virus can survive outside its living hosts. Therefore, virus mutations would occur more easily than we perceived.
In the past, we studied the adaptation of influenza A viruses to temperature changes with respect to their six proteins, hemagglutinin, neuraminidase, nucleoprotein, polymerase acidic protein, polymerase basic protein 2, and matrix protein 2 [8-13]. However, there are ten or eleven different proteins in influenza A viruses according to different strains, so it is necessary to study the adaptations in the remaining proteins from influenza A viruses in order to get a whole picture. Moreover, a considerable amount of new data concerning influenza A virus as well as temperature just become available, so it is necessary to incorporate these new data into this type of analysis.
The polymerase basic protein 1 (PB1) is a protein from influenza A virus as a subunit of RNA-dependent RNA polymerase complex. It associates with the transcription and replication of the influenza A viral genome [14], so it is important for the efficient propagation of the virus in the host and for its adaptation to new hosts [15]. As a part of influenza A virus, PB1 needs to adapt to any change in its host and its adaptation could lay the foundation for the survival of influenza A virus, of which new strain could lead epidemic as well as pandemic. Nowadays it is considered as a determinant of the pathogenicity of the 1918 pandemic virus [16]. Also, influenza PB1 is the major target for both CD4(+) and CD8(+) T-cell responses [17]. In this study, we analyzed the adaptation of PB1 proteins to temperature change.
2. MATERIALS AND METHODS
2.1. Temperature Data
The global temperature anomalies from 1850 to 2011, whose anomaly was based on the period 1961-1990, were obtained from HadCRUT3v data set that was the variance adjusted version of combined land and marine temperature anomalies on a 5˚ by 5˚ grid-box basis and from CRUTEM4v data set that was the variance adjusted version of land air temperature anomalies on a 5˚ by 5˚ grid-box basis [18,19]. The local temperature from 1956 to 1998 based on 0.5˚ by 0.5˚ latitude and longitude gridbox basis cross globe was obtained from New et al. [20].
2.2. PB1 Data
10732 full-length PB1 proteins of influenza A viruses sampled from 1956-2011 were obtained from the Influenza Virus Resources [21]. All groups of identical sequences in the dataset were represented by the oldest sequence in the group so that 3841 PB1 proteins were actually used in this study.
2.3. Adaptation in PB1 to Temperature Change
Virus adaptation is a mutation process along the time course, and interestingly temperature change is a process along the time course, too. These two facts dictate that we can plot both temperature change and virus adaptation along the time course, observe their relationship and determine whether virus adaptation occurs to suit temperature change.
At the level of species, adaptation can refer to features in population of species, for example, the number of population changes along the time course. At the level of individual in species, adaptation can refer to features in an individual in species, for example, the individual’s bodyweight changes along the time course. At the level of protein, adaptation can refer to features in a protein family, for example, the amino acid composition changes in protein family along the time course. In agreement with our previous studies along this line [8-13], we used the amino-acid pair predictability as a measure to present PB1’s adaptation along the time course [22].
For example, a PB1 protein (accession number BAD- 89303) has 757 amino acids, which are counted as 756 adjacent amino-acid pairs. There are 49 arginines “R” and 47 isoleucines “I” in this protein. If permutation can predict the appearance of amino-acid pair RI, then it would appear 3 times (49/757 × 47/756 × 756 = 3.04). Actually it does appear 3 times, so its appearance is predictable. By clear contrast, there were 50 glutamic acids “E” and 48 serines “S” in this PB1 protein. According to the permutation, the amino-acid pair ES would appear 3 times (50/757 × 48/756 × 756 = 3.17), however, it appeared 7 times in realty, which was unpredictable. In this way, we classified all amino-acid pairs in this protein as predictable and unpredictable portions, which are 28.04% and 71.96%. For another example, another PB1 protein (accession number BAD89323) has only one amino acid at position 80 different from BAD89303 PB1 protein, and its predictable and unpredictable portions are 27.38% and 72.62%.
In this manner, each PB1 protein has its own predictable/unpredictable portion, and 3841 PB1 proteins have 3841 different predictable/unpredictable portions. As each PB1 protein had its sampling year, so we can plot either predictable portion or unpredictable portion as a number along the time course to compare with the temperature trend over time.
2.4. Statistics
Linear regression was used to correlate the relationship between temperate anomaly and time and between unpredictable portion and time. In this way, the regressed lines can be used to observe the trends, which should be considered as the first stage of studies, because more sophisticated methods should be applied after obtaining the confirmative results from the first stage.
3. RESULTS
Figure 1 shows the similar trends in temperate change and in PB1 adaptation, which represents as averaged unpredictable portion for each year. For example, there are 3 PB1 proteins in 1956 and 53 PB1s in 2008. Statistically, the data are unbalanced because of difficulties in collection of samples. Consequently this would partially contribute to the difference between three regressions in upper, middle and lower panels.
Still, we can use the point-to-point method to furthermore analyze the trends in Figure 1, that is, we couple each PB1 protein with the temperature recorded at the same place and at same year, where and when that PB1 was sampled. For example, a PB1 protein (accession number AAA43641) was sampled in 1956 at Beijing, whose latitude and longitude are 39.91˚ and 116.40˚ according to Get Lat Lon [23], where the average yearly temperature was 10.93˚C in 1956 according to the 0.5˚ by 0.5˚ latitude and longitude grid-box basis cross globe obtained from New et al. [20].
Figure 2 displays 787 point-to-point relationships be-