Cell-Free DNA Extraction of Vitreous and Aqueous Humor Specimens for Diagnosis and Monitoring of Vitreoretinal Lymphoma

Summary

A procedure to extract cell-free DNA from vitreous and aqueous humor to perform molecular studies for diagnosing vitreoretinal lymphoma is established here. The method offers the ability to concurrently extract DNA from the cellular component of the sample or to reserve it for ancillary testing.

Abstract

Vitreoretinal lymphoma (VRL) represents an aggressive lymphoma, often categorized as primary central nervous system diffuse large B-cell lymphoma. To diagnose VRL, specimens such as vitreous humor and, more recently, aqueous humor are collected. Diagnostic testing for VRL on these specimens includes cytology, flow cytometry, and molecular testing. However, both cytopathology and flow cytometry, along with molecular testing using cellular DNA, necessitate intact whole cells. The challenge lies in the fact that vitreous and aqueous humor typically have low cellularity, and many cells get destroyed during collection, storage, and processing. Moreover, these specimens pose additional difficulties for molecular testing due to the high viscosity of vitreous humor and the low volume of both vitreous and aqueous humor. This study proposes a method for extracting cell-free DNA from vitreous and aqueous specimens. This approach complements the extraction of cellular DNA or allows the cellular component of these specimens to be utilized for other diagnostic methods, including cytology and flow cytometry.

Introduction

Vitreoretinal lymphoma (VRL) is an aggressive lymphoma associated with primary central nervous system diffuse large B-cell lymphoma^1,2,3. VRL is typically fatal due to its involvement in the central nervous system^1,2. Although rare^1,4, VRL often presents with symptoms similar to posterior uveitis and other vitreoretinal diseases^4,5. Consequently, patients exhibiting uveitis symptoms require a diagnosis to either confirm or rule out VRL.

Recently, consensus criteria for diagnosing VRL were published, which involve a combination of clinical examination and laboratory findings⁶. Specimens commonly used to diagnose VRL include vitreous humor and, more recently, aqueous humor⁷. Vitreous humor is obtained through a surgical procedure called pars plana vitrectomy, which allows access to the posterior segment of the eye⁸.

In the presented protocol, both aqueous humor and vitreous humor specimens were collected for cellular and cfDNA extraction. After anesthetizing the patients and placing trocars approximately 4 mm from the corneal limbus, an aqueous humor sample of approximately 100-200 µL was obtained using a 1 mL tuberculin syringe at the corneal limbus. For pseudophakic patients, undiluted vitreous was obtained by introducing sterile air into the infusion, enabling the collection of a larger amount of undiluted vitreous (up to 3.5 mL). In phakic patients, approximately 500 to 1000 µL of undiluted vitreous was removed before turning on an infusion of balanced salt solution. In some cases, secondarily diluted vitreous (500 to 2,000 µL) was collected by switching the infusion to fluid and placing the vitrector within the vitreous skirt to obtain this sample. The most dilute vitreous fraction was collected by preserving the cassette bag (Supplemental Figure 1) at the end of the surgery. Once this bag reached the pathology department, dilute vitreous was obtained from draining fluid out of this bag into conical tubes for subsequent DNA extraction.

Cytopathology of vitreous fluid is often considered the gold standard⁹. However, several studies have demonstrated limited sensitivity due to processing and minimal cellularity^10,11,12. Flow cytometry can aid in identifying clonal B-cells but can also be limited by low cellularity and the fragility of large lymphoma cells^13,14,15. Both cytopathology and flow cytometry requires intact whole cells. Many of these cells are destroyed during collection, storage, and processing. When molecular testing is performed using DNA extracted from intact cells (cellular DNA), it suffers from this same limitation. In addition, dividing the limited vitreous specimen for all of these tests reduces the amount of material available for each test.

Cell-free DNA (cfDNA) represents another source of DNA that does not require intact cells. cfDNA from vitreous specimens has been used for the detection of VRL^16,17 as well as uveal melanoma¹⁸. In this protocol, cellular and cell-free DNA are extracted from vitreous and aqueous fluid to detect VRL.

Protocol

The present protocol follows the human care guidelines and with approval of the institutional review board (IRB) of the University of Michigan. A waiver of informed consent was obtained for this from IRB. There are no relevant inclusion or exclusion criteria for the patients involved.

1. Separation of cellular and cell-free components

NOTE: Three types of samples can be received for VRL diagnostic testing: vitreous (undilute; fluid from vitrectomy collected prior to beginning infusion), diluted vitreous within a cassette bag (Figure 1), and aqueous humor (fluid from the anterior chamber of the eye).

1. For undiluted vitreous fluid, which is very viscous, dilute with 3-5 mL PBS to facilitate pipetting.
2. For dilute vitreous within the cassette bag (Supplemental Figure 1), drain fluid into 50 mL conical tubes (one for every 45 mL of fluid).
3. Aqueous humor is very low in volume (<200 µL). To ensure maximum recovery of cells, transfer the sample to a 1.5 mL tube, rinse the original 500 µL microcentrifuge tube the sample was collected in with an equal volume of PBS, and add to the 1.5 mL tube.
4. Centrifuge vitreous or aqueous fluid at 3,000 x g for 15 min at room temperature.
5. Carefully remove the supernatant using a pipette, being careful not to disturb the pelleted cells.

2. DNA extraction from the cellular component

1. Add an appropriate amount of cell lysis solution (300 µL for small pellet, 1 mL or more for larger pellet) (see Table of Materials). Mix by pipetting up and down.
2. When completely lysed, add 1/3 total volume of the protein precipitate solution (100 µL for 300 µL lysate) (see Table of Materials).
3. Vortex vigorously for 20-30 s. Spin for 3 min at 3,000 x g at room temperature.
4. Pipette 300 µL of isopropanol into a clean microcentrifuge tube. Carefully pipette the supernatant from the previous step to the isopropanol tube.
5. Add 1.5 µL of 20 mg/mL glycogen (for 300 µL lysis) (see Table of Materials).
6. Mix well by inversion, then place the sample in a -20 °C freezer for 1 h to overnight to aid in DNA precipitation.
7. Centrifuge the 1.5 mL tube for 5 min at 3,000 x g at room temperature. Pipette off the isopropanol.
8. Add 0.5 mL of 70% ethanol and invert the tube to wash the pellet.
9. Spin for 1 min at 3,000 x g at room temperature and decant the ethanol. Repeat the ethanol wash once.
10. Invert the tube on clean gauze to drain the solution and dry the pellet or decant. Quickly spin the DNA in a microcentrifuge and pipette the remaining drop of ethanol with a fine-tip disposable pipet tip (once no residual ethanol is left in the tube, the DNA should be dry and ready to hydrate in 5-10 min).
11. Add 45 µL of DNA hydration solution (see Table of Materials) to the dried DNA pellet.
12. Place the tube in a heating block at 50 °C and allow it to solubilize for 1-2 h. Pipet the DNA up and down to speed up the hydration process.
    ​NOTE: Add more DNA hydration solution if the sample appears too viscous.

3. Cell-free DNA extraction

1. The amount of supernatant from step 1.5 is variable. Add 70 µL of conditioning buffer (see Table of Materials) for every milliliter of the supernatant.
2. Mix clearing beads well by vortexing. Add 10 µL of clearing beads if processing <14 mL of cell-free supernatant, 20 µL if processing 14-40 mL.
3. Mix sample/bead mixture well by vortexing. Centrifuge at 3,000 x g for 15 min at room temperature.
4. Without disturbing the pellet, pipette out the supernatant leaving behind 100 µL if 10 µL of clearing beads are used, and 200 µL if 20 µL of clearing beads are used.
5. Add an equal volume of Urine Pellet Digestion buffer (see Table of Materials) to the pellet and resuspend the pellet well by vortexing or pipetting.
6. Add Proteinase K (5% (v/v)) to the resuspended pellet (e.g., add 10 µL of Proteinase K to 200 µL mixture) and mix well by gentle vortexing.
7. Incubate the pellet mixture at 55 °C for 30 min.
8. Add one volume of Genomic lysis buffer (see Table of Materials) to the digestion mixture (e.g., 210 µL Genomic lysis buffer to 210 µL of digestion mixture) and mix by vortexing.
9. Transfer the commercially available spin column (see Table of Materials) to a new collection tube.
10. Add 200 µL of Urine DNA Prep buffer to the spin column, centrifuge at ≥16,000 x g for 1 min at room temperature, and discard flow-through.
11. Add 700 µL of Urine DNA wash buffer to the column, centrifuge at ≥16,000 x g for 1 min at room temperature and discard flow-through.
12. Repeat the step with 200 µL Urine DNA wash buffer.
13. Transfer the spin column to a DNase/RNase-free microcentrifuge tube.
14. Add the appropriate volume of DNA elution buffer (based on the molecular tests to be performed) directly onto the column matrix and allow it to stand for 3-5 min at room temperature.
    NOTE: 30 µL of elution buffer was used in the example cases. More elution buffer is used if additional testing is required.
15. Centrifuge at ≥16,000 x g at room temperature for 1 min to obtain the cell-free DNA.

Representative Results

These extraction methods were performed on a limited number of cases to ensure adequate yield and amplifiability of cell-free DNA in comparison with cellular DNA from the vitreous (four) and aqueous (four) specimens. From these samples, DNA yields from the cell-free component of these fluids are similar to that of the cellular component (Table 1). Cellular and cfDNA from these samples were also evaluated using molecular testing, such as for the VRL-associated mutation MYD88 L265P. An allele-specific real-time PCR for MYD88 L265P (Figure 1) illustrates the detection of this VRL-associated mutation in both cellular DNA and cfDNA. In this example, the burden of disease appears to be higher in the cell-free component, as illustrated by a lower (cycle threshold (Ct). These data illustrate that cfDNA extracted from vitreous and aqueous humor using this method results in a source of DNA that can also be applied to diagnostic molecular testing.

Figure 1: MYD88 L265P allele-specific real-time PCR. Representative amplification plot showing MYD88 L265P allele-specific real-time PCR for cellular and cell-free DNA (each PCR reaction performed in duplicate). The threshold for cycle threshold (Ct) is indicated by the green line. The points at which amplification of cfDNA (blue and brown traces) and cellular DNA (red and green traces) - each performed in duplicate - reach the threshold are labeled. Background fluorescence is present near 1.00e-003. A higher mutation burden is present in the cell-free component, as demonstrated by a lower Ct. Please click here to view a larger version of this figure.

	Cellular	Cell-Free
Vitreous 1	130.5	67.5
Vitreous 2	337.5	105
Vitreous 3	150	168
Vitreous 4	1816	654
Aqueous 1	58.5	40.5
Aqueous 2	153	225
Aqueous 3	117	454.5
Aqueous 4	TL	27

Table 1: Cellular and cell-free DNA extraction yields. DNA yield (in ng) for four undiluted vitreous and four aqueous humor (anterior chamber) fluids from four different individuals based on fluorometric DNA quantitation. TL = too low to quantitate.

Supplemental Figure 1: Vitreous cassette bag. Example of vitreous cassette bag containing dilute vitreous fluid. Please click here to download this File.

Discussion

Vitreoretinal lymphoma (VRL) is an aggressive large B-cell lymphoma^1,2,3 whose symptoms can mimic other vitreoretinal diseases^4,5. Molecular testing of vitreous and, more recently, aqueous humor has become a critical method for making the diagnosis of VRL or ruling it out. However, these fluids are very low in volume and often have low cellularity. Many of the cells within these fluids can also be damaged during collection, storage, and processing^13,14,15. As a result, DNA yields from these specimens are often low. In addition, the cells from these fluids must be shared with other diagnostic methods, including cytopathology and flow cytometry. Cell-free DNA (cfDNA) within these fluids provides another source of DNA that does not require intact cells.

This protocol separates the cellular and cell-free components of vitreous or aqueous fluid. Vitreous fluid is diluted with PBS to enable pipetting due to its viscosity. Aqueous fluid is handled to ensure maximal recovery of this very low-volume fluid.

A comparison of DNA yields of these two components illustrates that substantial additional nucleic acid can be derived from the cell-free component. Molecular testing of cfDNA, such as MYD88 L265P allele-specific real-time PCR, demonstrates that VRL can be detected in the cell-free component of vitreous and aqueous humor as well as the cellular component. In many cases, the relative amount of VRL is higher in the cell-free component than in the cellular (Figure 1).

Extraction of the cell-free component of vitreous and aqueous fluid enables molecular evaluation of a component of these fluids that would otherwise be discarded. Because VRL is represented in cellular and cell-free components, cellular DNA and cfDNA could be combined to maximize the available DNA from these limited specimens. Alternatively, the cell-free component of these fluids could be used for molecular testing, and the cellular component could be used for testing that requires intact cells, i.e., cytopathology and flow cytometry. This approach would avoid separate aliquots of these fluids being allocated to each test which can compromise the sensitivity of each test for the detection of VRL. Given the success of this method in vitreous and aqueous specimens, this method may also be useful in other limited fluids to increase the amount of DNA available for molecular testing.

Materials

Name	Company	Catalog Number	Comments
2-Propanol (Isopropanol)	Fischer	A415-500
DNA Clean & Concentrator-10	Zymo Research	D4011
DNA Clean & Concentrator-5	Zymo Research	D4003
Gentra Puregene Cell Lysis Solution	Qiagen	158906
Gentra Puregene DNA Hydration Solution	Qiagen	158916
Gentra Puregene Protein Precipitation Solution	Qiagen	158912
Phosphate Buffered Saline (PBS)	Sigma	P-4417
Quick-DNA Urine Kit	Zymo Research	D3061	Conditioning buffer; also includes clearing beads, Proteinase K and spin columns
Ultrapure Glycogen, 20 µg/µL (20 mg/mL)	Thermo Fisher/Invitrogen	10814010