Effect of EBV-Transformation on Oxidative Phosphorylation Physiology in Human Cell lines

Do the immortalized and cryopreserved white blood cells that are part of the 1,000 Human Genomes Project represent a valuable cellular physiological resource to investigate the importance of genome wide sequence variation? While much research exists on the nucleotide variation in the 1,000 Human Genomes, there are few quantitative measures of these humans’ physiologies. Fortunately, physiological measures can be done on the immortalized and preserved cells from each of the more than 1,000 individuals that are part of Human Genome project. However, these human white blood cells were immortalized by transforming them with the Epstein-Barr virus (EBV-transformed lymphoblastoid cell lines (LCL)). This transformation integrates the viral genome into the human genome and potentially affects important biological differences among individuals. The questions we address here are whether EBV transformations significantly alters the cellular physiology so that 1) replicate transformations within an individual are significantly different, and 2) whether the variance among replicates obscures the variation among individuals. To address these questions, we quantified oxidative phosphorylation (OxPhos) metabolism in LCLs from six individuals with 4 separate and independent EBV-transformations. We examined OxPhos because it is critical for energy production, and mutations in this pathway are responsible for most inborn metabolic diseases. The data presented here demonstrate that there are small but significant effects of EBV-transformations on some OxPhos parameters. In spite of significant variation due to transformations, there is greater and significant variation among individuals in their OxPhos metabolism. Thus, the LCLs from the 1,000 Human Genome project could provide valuable insights into the natural variation of cellular physiology because there is statistically significant variation among individuals when using these EBV-transformed cells.


ABSTRACT:
Do the immortalized and cryopreserved white blood cells that are part of the 1,000 Human Genomes Project represent a valuable cellular physiological resource to investigate the importance of genome wide sequence variation? While much research exists on the nucleotide variation in the 1,000 Human Genomes, there are few quantitative measures of these humans' physiologies. Fortunately, physiological measures can be done on the immortalized and preserved cells from each of the more than 1,000 individuals that are part of Human Genome project. However, these human white blood cells were immortalized by transforming them with the Epstein-Barr virus (EBV-transformed lymphoblastoid cell lines (LCL)). This transformation integrates the viral genome into the human genome and potentially affects important biological differences among individuals. The questions we address here are whether EBV transformations significantly alters the cellular physiology so that 1) replicate transformations within an individual are significantly different, and 2) whether the variance among replicates obscures the variation among individuals. To address these questions, we quantified oxidative phosphorylation (OxPhos) metabolism in LCLs from six individuals with 4 separate and independent EBV-transformations. We examined OxPhos because it is critical for energy production, and mutations in this pathway are responsible for most inborn metabolic diseases. The data presented here demonstrate that there are small but significant effects of EBV-transformations on some OxPhos parameters. In spite of significant variation due to transformations, there is greater and significant variation among individuals in their OxPhos metabolism. Thus, the LCLs from the 1,000 Human Genome project could provide valuable insights into the natural variation of cellular physiology because there is statistically significant variation among individuals when using these EBV-transformed cells.
Many mitochondrial diseases are caused by point mutations in one of the 13 mitochondrial encoded proteins or mutations in the mt-tRNAs [4,8,9]. Most of these mtDNA mutations have an adult disease onset and can have very specific cellular targets (e.g., lesions in a specific area of the basal ganglion [4,8,9]). Along with the 13 mtDNA genes, 76 nuclear genes encode proteins for the 5 OxPhos enzyme complexes ( Figure 1).
In contrast to mtDNA mutations, deleterious mutations in these nuclear genes affect more individuals, have greater penetrance, are typically lethal, and have early onset (within the first year of life; [4][5][6]). Surprisingly, mtDNA mutations affecting OxPhos proteins have a heterogeneous disease presentation and variable penetrance even among homoplasmic individuals [2,4,5], and many individuals have nuclear SNPs associated with pathogenic diseases but remain healthy [10][11][12][13]. These observations suggest that interactions among DNA polymorphisms in both mitochondrial and nuclear genomes have significant effects on human health by affecting the OxPhos pathway. To better understand these observations would involve quantifying the natural variation in OxPhos metabolism taking into consideration genomic polymorphisms.
Quantitative analyses of OxPhos metabolism with knowledge about DNA sequence variation in OxPhos genes and the approximately 1,5000 other nuclear genes that affect mitochondrial ATP production [1] is possible using the resources from the "1,000 Figure 1. OxPhos pathway. OxPhos consists of five enzyme complexes with both mitochondrial and nuclear genomes encode proteins in four of these complexes. Numbers of mitochondrial and nuclear encoded proteins in the different subunits are listed. http://www.genome.jp/kegg-bin/show_pathway?map00190 [31][32][33]  Human Genome" project [14]. The individuals in the 1,000 Human Genome collection represent much of the genetic diversity in humans worldwide, with sampled populations ranging from East and South Asia, Europe, West Africa, and the Americas [14].
Importantly, in addition to having fully sequenced genomes, these individuals have cryopreserved, immortalized, lymphoblastoid cell lines (LCL). These LCLs originate as peripheral B lymphocytes that have been isolated from human subjects and transformed with the Epstein-Barr virus (EBV) [15].
Transformed LCLs are effectively immortal, express a wide array of genes including many metabolic pathways, and are easy to culture in a laboratory setting [15].
EBV-transformed LCLs and specifically cell lines from the 1,000 Human genome project have been used successfully to define sex bias in gene expression [16], statin-dependent QTL [17], cell specific networks [18], genetic variation in DNA replication [19] and epigenetics [20,21]. Overall, these and other studies have concluded that EBVtransformed LCLs continue to represent naturally occurring biological variation [22,23]. However, EBV-transformation can have significant effects on gene expression, and there is considerable debate on whether this transformation alters biologically important phenotypic differences, including cellular physiology, which would reduce the EBV-transformed LCLs' utility for understanding the effects of nucleotide variation [20,24,25]. Because it is unclear if the random viral integration affects physiological variation in LCLs, we sought to quantify the variation that arises from this transformation by quantifying the metabolic activity of the oxidative phosphorylation (OxPhos) pathway. This approach would have been strengthened by comparing pretransformed with the currently available transformed LCLs, but these untransformed cells are unavailable. Instead, presented here we investigated whether independently transformed LCLs within individuals introduces significantly large amount of variance such that there is little remaining variation among individuals (i.e. variation within individual is greater than the variation among individuals).
To investigate the effect of EBV-transformation, we used six individual donors each with four, independent, replicate EBV-transformed cell lines [26]. These independent EBV-transformations are used to quantify the effect of transformation and whether EBV-transformation obscured significant variation among individuals. For our study we focused on OxPhos metabolism, and we measured the six parameters of the OxPhos pathway: State 3 activity, E State activity, Complex I activity, Complex II activity, Complex IV activity, and proton leak across the mitochondrial membrane.
Overall, EBV-transformation had a small but significant effect, yet this effect did not obscure differences among individuals.  in upright T25 tissue culture flasks and kept at 37°C, 70% humidity, and 5% carbon dioxide. Cell lines were cultured for 2 weeks before OxPhos measurements to ensure that they had sufficient acclimation in the provided media. Cells were diluted approximately every 3 days between 30-70% to achieve 300,000 cells/mL. The day prior to OxPhos determination cells were diluted to 500,000 cells/mL, and the cells would grow to approximately 1,000,000 cells/mL on the day of measurement. 1.5 million cells were used to measure OxPhos metabolism.

Cell Culturing
The cell concentration in a given culture was estimated by staining a sample with a 2:1 ratio of cells to Trypan blue dye for 3-5 minutes. The cells were then counted on a glass haemocytometer. DNA concentration was determined and used as the measure of cell density for each LCL sample used for OxPhos measurements.

Oxidative Phosphorylation Measurements
The OxPhos metabolic parameters are defined in Table 2

Statistical Analyses
The normal distribution of all six OxPhos parameters (   (Table 1).    Table 4). Among these three individuals, the significant effects of transformation were from OxPhos measured in different weeks. That is, although all six individuals were measured in triplicate weekly, the separate transformations were measured in different  (Table 2), and order of significance were different for different individuals ( where the first assay is the highest and the fourth assay is the lowest.

EBV-transformation has had
varied effects on LCLs. In the one of the first studies [20], EBVtransformation relative to whole blood had small but significant effects on mRNA expression and methylation patterns, yet LCLs recapitulate the naturally occurring gene expression variation in primary B cells [20]. Similarly in EBV-transformed LCLs cells, cell cycle genes have significantly greater expression compared to primary tissues, which is thought to be the result of less repressive transcription factor regulation [18]. These differences paired LCLs with whole blood and found that the regulatory changes were subtle; for example the levels of mRNA expression for the transcription factor Sp1 and Sp3 were not significantly different in transformed versus native cells, but the targets for these transcription factors differed [18]. Furthermore, mRNA expression substantially changed with passage [16,18,25]. In contrast, EBV-transformation had no significant effect on mitochondrial copy number, with nearly equal numbers of mitochondria among replicate EBV-transformations [20]. Additionally, EBVtransformation did not obscure sex differences among 99 individuals [16]. Similarly, EBV-transformation did not obscure the association of 16 regulator genes with DNA replications among 161 individual from the EBV-transformed LCLs in 1,000 Human Genome collection [19]. Furthermore, EBV-transform cells are used regularly in pharmacological studies with great success [17,23]. These data, while acknowledging EBV-transformation effects, find that transformation effects do not inhibit genetics studies. What is unclear is whether these EBV-transformation effects on molecular and biochemical traits might affect more complex physiological processes.
In the study presented here, we compared six individuals with four separate and independent EBV-transformations to inquire if EBV-transformation significantly Table 4. EBV Significant Effect. The three individuals among the six with significant EBV effects. Numbers refer to the order and transformation that was significant (see Table 2 study, not making a type II error is more important because our concern is that EBVtransformations significantly alters complex physiological phenotypes. Thus, the uncorrected statistical tests are most relevant.
The data presented here suggest that EBV-transformation can affect OxPhos metabolism but only in some OxPhos traits for some individuals. Further, this effect is not large enough to obscure the differences among individuals. Overall, the insight gained from this study provides greater understanding of how EBV integration into the genome of LCLs can influence phenotypes such as the metabolic activity of the OxPhos pathway. It also enhances our knowledge of OxPhos metabolic variation in LCLs, which will be essential for future research in this area of study. Most importantly, these findings indicate that we can effectively study OxPhos metabolism as a phenotype among different EBV-transformed LCLs and compare OxPhos to mRNA expression, nuclear genome sequence variation, and mitochondrial genome sequence variation.
These types of studies would enhance our understanding of the genetic variation found in the 1,000 Human Genome collection.

DATA AVAILABILITY
Raw, and normalize (rates/[DNA]) for all Oxphos measurements for all individuals will be made available in Dyrad. XXYY.