Viral Load Measurements for Kaposi Sarcoma Herpesvirus (KSHV/HHV8): Review and an Updated Assay

Abstract

"When you can measure what you are speaking about, and express it in numbers, you know something about it." is a famous quote attributed to Lord Kelvin. This sentiment puts viral load measurements at the center of virology. Viral load, or more precisely, DNA copy number measurements, are also used to follow infections with human herpesviruses, such as Kaposi sarcoma herpesvirus (KSHV) and Epstein-Barr Virus (EBV). EBV and KSHV are associated with human cancers, and determining their DNA copy numbers in the context of cancer prediction and progression on therapy is of fundamental scientific and translational interest. Yet, there is no generally accepted assay for KSHV DNA quantitation, and KSHV viral load is not used in clinical decision-making. Here, we review the history of KSHV DNA detection assays, explore factors that affect sensitivity and specificity, and describe an automated, high-throughput, real-time quantitative polymerase chain reaction (PCR) assay for KSHV and EBV. In conjunction with a digital PCR assay using the same primer/probe combination, we describe how to determine the absolute KSHV genome copy numbers in plasma, peripheral blood mononuclear cells, saliva, and other easily accessible body fluids.

Keywords: Castleman's disease; EBV; KSHV; Kaposi sarcoma; MCD; PEL; lymphoma; real‐time QPCR.

Figure 1:. Recovery of input virus by assay volume and instrument accuracy.

(A) Box and whisker plot comparing the sensitivity of the smaller volume (18.9 μL) assay in red with the larger volume (94.5 μL) in blue for two primer pairs EBV and KSHV. The starting point in each case is the same 1 mL sample, i.e., a fixed amount of virus, processed through DNA isolation and eluted in 100 μL. The vertical axis shows the number of copies recovered. The bolded horizontal line within the box plot indicates the median of the data set. The box itself represents the interquartile range (IQR). The ends of the whiskers represent the minimum and maximum values that are not considered outliers. Outliers are not shown. The absolute limit of detection was set to 100 cps/ml (dash-dotted line). (B) Box and whisker plot comparing the sensitivity of the small (CAS-9) liquid handling robot, in green, versus the large (Tecan) liquid handling robot, in black, using the same small (18.9 μL) volume assay. The vertical axis shows the number of copies recovered.

Figure 2:. PEG-based Enrichment.

Average copy number recovered in samples incubated with PEG (Blue) versus samples with no incubation in (Red) for the same input. Each data set on the horizontal axis, A-C, represents a separate run with different amounts of input virus. The vertical axis shows the number of copies recovered. The bolded horizontal line within the box plot indicates the median of the data set. The box itself represents the interquartile range (IQR). The ends of the whiskers represent the minimum and maximum values that are not considered outliers. Outliers are not shown.

Figure 3. Endpoint PCR analysis.

(A) Automated electrophoresis was conducted on QPCR products from three virally infected human cell lines: BCBL-1, Namalwa (Nam), and BC-1. Also shown is the non-template control (NTC). Amplicon size is measured in base pairs (bp) with 1500 bp as the upper limit marker (UM) and 25 bp as the lower limit marker (LM). (B) Microfluidics-based electrophoresis of a ~100 bp EBV amplicon (32 run time/s) was shown to elute after our smaller 70 bp oligo control amplicon (28 run time/s). Excess primer or primer dimer is around 22 run time/s by the lower limit (LM). “UM” indicates the upper size limit that the gel matrix can resolve. (C) Correspondence between melting temperature and amplicon size as measured by microfluidics-based electrophoresis. “a.” points to the QPCR amplification product from the sample (“s.1,” “s.2”) or positive control (“pos”), “b.” to that of the control oligo (“O), “c.” to excess primer. No products are seen in the NTC lanes. Relative molecular weights (“mw”) are shown in lane 1. “UM” and “LM” refer to the upper and lower resolving limits of the gel matrix, respectively.

Figure 4. Examples of a spike-in control and batch normalization.

(A) This is a gray violin plot with the data points overlayed as red dots. The raw Cp values are shown on the vertical axis, and random QPCR plate identifiers are shown on the horizontal axis. Each plate contains 24 samples. The black arrows indicate samples for plates 717 (n=1) and 735 (n = 2), where the FLY extraction control failed. (B) Batch normalization across multiple PCR plates using Cp’ values The raw Cp values are shown on the horizontal axis, and the batch-normalized Cp’ values are on the vertical axis. For each plate (indicated by different colors), at least two data points, typically four, are used to calculate a standard curve.

Figure 5:. Elimination of Non-Virion DNA.

(A) Three concentrations of DNase (Promega) or mock control were added to samples that were concentrated with PEG. (B) A density plot of 80 primer pairs that were designed to span the entire length of the viral genome. KSHV (top), but not human DNA (bottom), is protected from DNase I digestion. (C) Individual Primer pair performance and non-template control (NTC). KSHV-positive human specimens were treated with DNase (blue) or mock (red). Technical triplicates for all primers were measured using a standard QPCR detection level (40-CT).

Figure 6.. Correspondence of real-time QPCR and digital PCR.

Shown are the DNA copy numbers per μL as ascertained by digital PCR on the vertical axis (in log 10 scale) and the raw Cp values for the same sample on the horizontal axis for (A) EBV, (B) ERV, and (C) KSHV as well as clinical samples (D24). Also shown are the linear regression lines and ^95%CI in gray. The regression line is calculated based on cell lines with known cell numbers.