use of a flow-cell system to investigate virucidal dimethylmethylene blue phototreatment in two...

B L O O D C O M P O N E N T S

Use of a flow-cell system to investigatevirucidal dimethylmethylene blue

phototreatment in two RBC additive solutions

Stephen Wagner, Andrey Skripchenko, and Dedeene Thompson-Montgomery

BACKGROUND: Limited photoinactivation kinetics, useof low-volume 30 percent Hct RBCs, and hemolysishave restricted the practicality of the use of dimethyl-methylene blue (DMMB) and light for RBC decontami-nation. A flow-cell system was developed to rapidly treatlarger volumes of oxygenated 45 percent Hct RBCs withhigh-intensity red light.MATERIALS AND METHODS: CPD-whole blood wasWBC reduced, RBCs were diluted in additive solutions(either Adsol or Erythrosol), and suspensions were sub-sequently oxygenated by gas overlay. Intracellular orextracellular VSV and DMMB were sequentially added.VSV-infected RBC suspensions (45% Hct) were passedthrough 1-mm-thick flow cells and illuminated. Sampleswere titered for VSV, stored for up to 42 days, and as-sayed for Hb, supernatant potassium, ATP, and MCV.RESULTS: The use of oxygenated RBCs resulted inrapid and reproducible photoinactivaton of �6.6 log ex-tracellular and approximately 4.0 log intracellular VSVindependent of additive solution. Phototreated AdsolRBCs exhibited more than 10 times greater hemolysisand 30 percent greater MCV during storage than identi-cally treated Erythrosol RBCs. Phototreatment causedRBC potassium leakage from RBCs in both additive so-lutions. ATP levels were better preserved in Erythrosolthan Adsol RBCs.CONCLUSION: A rapid, reproducible, and robustmethod for photoinactivating model virus in RBC sus-pensions was developed. Despite improved hemolysisand ATP levels in Erythrosol-phototreated RBCs, stor-age properties were not maintained for 42 days.

Careful donor selection and extensive labora-tory testing have greatly improved the safety ofthe US blood supply. Despite these highly suc-cessful measures, a small risk of viral transmis-

sion by transfusion still exists. Most transmission isthought to occur during the window period or before thedevelopment of detectable antigen, antibody, and/ornucleic acid by current test methods. The risk of viraltransmission per unit in the US has been estimated to be1 in 1,000,000 for HAV, 1 in 138,700 to 233,000 for HBV, 1in 250,000 to 2,000,000 for HTLV-I and -II, and 1 in 10,000for human parvovirus B19 and with the introduction ofNAT, 1 in 1,935 for HCV and 1 in 2,135 for HIV.1-3

Pathogen inactivation methods have been exploredin cellular blood components to further reduce the smallresidual transfusion-transmitted risks from tested virusesand to potentially lessen the risk from untested or unrec-ognized agents.4,5 The phenothiazine photosensitizer di-methylmethylene blue (DMMB) has been shown to inac-tivate lymphocytes and several model enveloped virusesin dilute RBC suspensions.6-8 Under virucidal conditions,RBC in vitro properties were adequately preserved during42-day 1 to 6�C storage.6 Despite these encouraging ini-tial data, some RBC membrane damage was observed,with approximately three times greater levels of hemoly-sis in phototreated samples than controls on Day 42 of

ABBREVIATIONS: DMMB = dimethylmethylene blue; LED =

light-emitting diode.

From the Jerome H. Holland Laboratory for the Biomedical

Sciences, Blood & Cell Therapy Development, American Red

Cross Biomedical Services, Rockville, Maryland.

Address reprint requests to: Stephen J. Wagner, PhD, Hol-

land Laboratory, 15601 Crabbs Branch Way, Rockville, MD

20855; e-mail: [email protected],

Supported in part by the National Heart Lung and Blood

Institute, Grant HL66779, and by a grant from Baxter Health-

care and Cerus Corporation.

Received for publication January 25, 2002; revision re-

ceived March 26, 2002, and accepted March 28, 2002.

TRANSFUSION 2002;42:1200-1205.

1200 TRANSFUSION Volume 42, September 2002

storage. In addition, inactivation of the single-strandedRNA model virus, VSV, was neither robust nor reproduc-ible, with 4.4 � 1.0 log extracellular virus inactivated un-der these conditions. Other concerns about the practical-ity of the method include the lengthy illumination time(20-30 min) and the use of low-Hct (30%), low-volume(4 mL) RBC suspensions in open systems.

This study investigates several modifications of thephotoinactivation system in an attempt to make the pro-cess more practical, reproducible, and robust. High-intensity red light-emitting diode (LED) sources and oxy-genated RBCs were used to ensure more reproducibleoxygen levels and accelerate photoinactivation kinetics.RBC suspensions were illuminated in a flow-cell systemwith the potential for treating entire units of RBCs in aclosed system. In addition, RBC storage properties andphotoinactivation efficiency by use of two different RBCadditive solutions were investigated.

MATERIALS AND METHODS

RBC preparation and oxygenationRBCs were prepared from units of whole blood (500 mL)collected in 70 mL CPD in triple-pack container systems(PL146, primary container, Baxter Healthcare, Deerfield,IL). Whole blood was WBC reduced via an in-line filter(Sepacell RZ-2000, Asahi Medical Co., Ltd, Tokyo, Japan),cooled to 1 to 6�C overnight, and subsequently centri-fuged at 1 to 6�C and 1471 � g for 4 minutes. The plasmasupernatant was expressed, and 110 mL of cold additivesolution (Adsol [Baxter Healthcare, Deerfield, IL] or Ery-throsol)9,10 was added to RBC. RBC suspensions in addi-tive solution were oxygenated by adding 240 mL of a 60 to40 percent O2 to N2 gas mixture to RBC suspensions in1-L containers (PL2410 plastic, Fenwal, Baxter Health-care) and by subsequent incubation for 30 minutes at 1 to6�C with agitation (orbital shaker, 100 r.p.m., 19-mm or-bit, VWR Scientific, West Chester, PA). The concentrationof O2 chosen for the O2 and N2 gas overlay was deter-mined empirically by measuring resulting blood gas lev-els. Oxygen levels were measured by use of a blood gasanalyzer (RapidLab 348, Bayer Corp., Medfield, MA) andwere routinely supersaturated with levels greater than400 mmHg.

Intracellular or extracellular VSV was added to oxy-genated RBC suspensions and thoroughly mixed. A 720�M DMMB stock solution in water was then added toAdsol or Erythrosol and mixed, and the diluted DMMBsolution was then added to oxygenated RBC suspensionsto yield a final concentration of 6 �M. The resulting sus-pension was incubated for 15 minutes at 1 to 6�C. Thesum of the volumes of virus and DMMB stock solutionswas less than 10 percent of the resulting 45 percent HctRBC suspension.

DMMB flow-cell systemDMMB was purchased (Aldrich Chemical Company, St.Louis, MO) and purified by a medium-pressure liquidchromatograph as previously described.6 Cold RBC sus-pensions containing DMMB were pumped through theflow system at a rate of 0.317 mL per second by use of avolumetric infusion pump (Flo-Gard 6200, Baxter-Travenol). Suspensions traveled from transfer packsthrough an irradiated and disposable flow cell (illumi-nated portion, 5.7 � 7.0 cm; PL2410 plastic, Fenwal) to areceiving transfer pack via standard blood bag tubing.The PL2410 plastic transmitted greater than 98 percent ofthe incident red light. The flexible flow cell was placedbetween two sheets of Plexiglas separated by spacers tocreate a blood film of approximately 1 mm thickness. Themean blood residence time within the flow cell was 13.1seconds. Two red LED sources (Q-beam 2001-MED,Quantum Devices, Inc., Barneveld, WI) illuminated eachPlexiglas sheet with 670 (peak intensity) � 13-nm (half-peak intensity) light with fluence rates adjustable up to7.16 mW per cm2. Fluence rates were measured by use ofa handheld laser power meter with a silicon cell sensor(Edmunds Industrial Optics, Barrington, NJ).

Extracellular VSV assayVSV was provided by Meg Lieu (Hyland Diagnostics, Du-arte, CA). We propagated VERO (isolated from Africangreen monkey kidney, CCL81, ATCC, Rockville, MD) cellsin medium (RPMI 1640 with glutamine, Biofluids, Rock-ville, MD) supplemented with 10 percent fetal bovine se-rum. Cells were seeded into six-well culture plates andallowed to grow to form confluent monolayers. Controland phototreated samples were serially diluted 1 in 10,plated onto confluent VERO cell monolayers, and incu-bated for 30 minutes with gentle rocking at 37�C for theadsorption of virus onto cells. The inoculum was aspi-rated and washed with PBS, a semiliquid agar layer (0.2%)was added to each well, and infected monolayers wereincubated at 37 � 2�C in air containing 5 percent CO2 for1 day. After incubation, the agar layer was removed byaspiration and the monolayer was stained with 0.1 per-cent crystal violet in ethanol for at least 15 minutes. Thestain was removed by aspiration, the plates were washedwith water, and the plaques were enumerated.

Intracellular VSV assayWe prepared virus-infected cells by inoculating a conflu-ent VERO monolayer with VSV at a virus-to-cell ratio ofgreater than 1. The inoculum was incubated for 30 min-utes at room temperature to allow entry of the virus andwas subsequently removed by aspiration. Infected mono-layers were washed with RPMI 1640 and detached fromthe culture flask by the addition of 0.05 percent trypsinand 0.02 percent versene solution for 5 minutes. The re-

FLOW SYSTEM FOR DMMB PHOTOTREATMENT

Volume 42, September 2002 TRANSFUSION 1201

sulting infected cell suspension was diluted approxi-mately 1 in 15 in RPMI 1640, and infected cells werecentrifuged at 200 � g for 10 minutes. The infected cellpellet was resuspended in additive solution and subse-quently added to WBC-reduced, oxygenated RBCs to givea final Hct of 45 percent. Control and phototreated RBCswere first diluted two times and then serially diluted tentimes. The twofold dilution was performed to eliminateRBC interference with VSV plaque formation. Infectedcells were inoculated onto confluent, uninfected VEROcell monolayers in six-well plates. As in the assay for ex-tracellular virus, semiliquid agar (0.2%) was added toeach well, and the infected monolayers were incubated at37�C in air containing 5 percent CO2 for 1 day. The agarlayer was removed by aspiration, and the monolayer waswashed with PBS to remove RBCs and stained by theaddition of 0.1 percent crystal violet in ethanol for at least15 minutes. After removal of the stain by aspiration,plates were washed with water. Plaques were counted,and they represented viral growth arising from intracel-lularly infected VERO cells.

RBC and supernatant assaysWe assayed RBCs for ATP by standard methods on fresh,unfrozen lysates (Technical Bulletins 226-UV, SigmaChemical Co.). Supernatant Hb was determined by thetetramethylbenzidine method (Procedure 527, Sigma).11

We measured extracellular potassium concentrations byuse of a blood gas analyzer (RapidLab 348, Bayer Corp).MCV and total Hb were measured by use of an automatedcell counter (Cell Dyn 3700, Abbott Laboratories, AbbottPark, IL).

Statistical analysisDetermination of means and SD of experimental valuesand performance of two-tailed t-tests were carried out byuse of standard software (Instat, GraphPad Software, SanDiego, CA).

RESULTS

VSV inactivationThe DMMB photoinactivation of extracellular VSV in 45percent Hct RBCs is given in Fig. 1. The inactivation ki-netics are psuedo-first order and comparable for RBCssuspended in Adsol and Erythrosol, with overlapping er-ror bars for all the Adsol and Erythrosol data points at agiven fluence. More than 6.6 log extracellular VSV is in-activated following treatment with 6 �M DMMB and 187mJ per cm2 red light.

The inactivation of intracellular VSV by with 6 �MDMMB and 187 mJ per cm2 red light in 45 percent HctRBC suspension was determined. Under these treatmentconditions, 4.03 � 0.59 and 4.04 � 0.29 log intracellular

VSV was inactivated in Adsol and Erythrosol suspensions,respectively.

RBC storage propertiesFigure 2 shows hemolysis of control and phototreated(6 �M DMMB and 187 mJ/cm2 red light) RBCs suspendedin Adsol and Erythrosol during 1 to 6�C storage. RBCs sus-pended in Adsol rapidly hemolyzed during storage, with

Fig. 1. Extracellular VSV inactivation kinetics following

DMMB phototreatment. RBC suspensions in Adsol (�, n = 3

units) or Erythrosol (�, n = 3 units) were subjected to 6 �M

DMMB and various fluences of red LED light. The dotted

line represents the sensitivity limit of the VSV plaque assay.

Fig. 2. Hemolysis during storage of RBCs phototreated with

DMMB. Control (�, �) and phototreated (�, �) RBC suspen-

sions in Adsol (�, �) or Erythrosol (�, �) were stored up to

42 days at 1 to 6�C and assayed for supernatant and total

Hb. Phototreated RBCs were subjected to 6 �M DMMB and

187 mJ per cm2 red LED light. �, n = 4 units; *, n = 5 units;

�, n = 26 units; �, n = 3 units. (Inset) The same data by use

of a different scale and emphasizes photoinduced hemolysis

from Erythrosol-containing samples.

WAGNER ET AL.


greater than 25 percent hemolysis on Day 42. With theexception of day 0 samples, all Adsol-phototreated unitshad greater than 1 percent hemolysis during storage. Incontrast, there was significantly less (p < 0.05) hemolysisfrom phototreated RBCs suspended in Erythrosol, with1.9 percent hemolysis on Day 42. Nevertheless, hemolysisfrom phototreated RBCs suspended in Erythrosol weresignificantly greater than those observed in untreatedcontrols (p < 0.05, Fig. 2, inset). For example, all pho-totreated Erythrosol units had greater than 1 percent he-molysis on Day 42, and two of three units had �1 percenthemolysis on Day 28. Control RBCs containing DMMBbut not illuminated and illuminated control RBCs lackingDMMB had similar hemolysis levels to those of untreatedcontrols (data not shown). Hemolysis of Erythrosol con-trol RBCs was comparable to that observed for AdsolRBCs.

ATP levels are given in Fig. 3 for control and pho-totreated (6 �M DMMB and 187 mJ/cm2 red light) RBCssuspended in Adsol or Erythrosol during 1 to 6�C storage.Phototreated Erythrosol RBCs had comparable (p > 0.05)initial and Day 42 levels to those from control RBCs, butless ATP than Erythrosol control cells on Day 28. In con-trast, phototreated Adsol RBCs had significantly less ATPlevels (p < 0.05) than those from Adsol or Erythrosol con-trols throughout storage.

Potassium efflux from phototreated and control cellssuspended in Adsol or Erythrosol during storage is givenin Fig. 4. Phototreated RBCs suspended in either Adsol orErythrosol rapidly released potassium to the supernatant,with levels five to seven times greater than those fromuntreated control cells (p << 0.05). In addition, potassium

levels in the supernatant from phototreated cells sus-pended in Erythrosol were consistently greater thanthose from identically treated cells suspended in Adsol(p < 0.05).

A comparison of the MCV of control and pho-totreated RBCs suspended in Adsol or Erythrosol is givenin Fig. 5. The MCV of control and phototreated RBCs

Fig. 3. ATP levels during storage of RBCs phototreated with

DMMB. RBC suspensions in Adsol or Erythrosol were stored

up to 42 days at 1 to 6�C and cell extracts were assayed for

ATP content. Phototreated RBCs were subjected to 6 �M

DMMB and 187 mJ per cm2 red LED light. Control Erythro-

sol RBCs (solid, n = 26 units); phototreated Erythrosol RBCs

(solid with vertical lines, n = 3 units); control Adsol RBCs

(open, n = 4 units); and phototreated Adsol RBCs (open with

diagonal lines, n = 5 units).

Fig. 4. RBC potassium efflux following DMMB phototreat-

ment. Control (�, �) and phototreated (�, �) RBC suspen-

sions in Adsol (�, �) or Erythrosol (�, �) were stored up to

7 days at 1 to 6�C and assayed for supernatant potassium.

Phototreated RBCs were subjected to 6 �M DMMB and 187

mJ per cm2 red LED light. �, n = 4 units; *, n = 5 units;

�, n = 26 units; �, n = 3 units.

Fig. 5. MCV following DMMB phototreatment. Control

(�, �, n = 5 units for each additive solution) and pho-

totreated (�, �, n = 5 units for each additive solution) RBC

suspensions in Adsol (�, �) or Erythrosol (�, �) were stored

up to 28 days at 1 to 6�C and MCV was measured. Pho-

totreated RBCs were subjected to 6 �M DMMB and 187 mJ

per cm2 red LED light.



remained relatively constant during storage of ErythrosolRBCs. The MCV of control Adsol RBCs also remainedfairly stable during storage. However, phototreated AdsolRBCs increased in MCV significantly during storage, withDay 28 values 30 percent greater than those measured onDay 0 (p < 0.05).

DISCUSSION

A flow-cell system was developed for DMMB photoinac-tivation of viruses in 45 percent Hct RBC suspensions.The use of oxygenated RBCs and high-intensity LEDsspecific for DMMB adsorption allowed rapid and repro-ducible photoinactivation of a model single-strandedRNA virus, with flow-cell residence times of 13.1 seconds.If desired, a more rapid throughput could be obtainedwith higher intensity light sources and/or with flow cellsand light sources of increased dimensions.

Despite similar virus inactivation kinetics, there werelarge differences in photoinduced hemolysis during stor-age, with levels in Adsol roughly 10 times greater thanthose observed in Erythrosol. Erythrosol differs from Ad-sol in that it is a hypoosmolar additive solution that lackssodium chloride, but provides inorganic phosphate andsodium citrate. Citrate ion is known to be impermeableto RBCs.12,13 Both storage solutions contain mannitol.Untreated control RBCs suspended in Erythrosol ap-peared to have less hemolysis on Day 42 (0.13 � 0.05%)than those observed in Adsol (0.24 � 012%), but thesedifferences were neither large nor significant.

It is generally thought that photodynamic hemolysisstems from colloidal osmotic shock.14 This shock is a re-sult of photooxidative damage to lipid and/or membraneproteins and leads to increases in ion permeability. Ion-permeable RBCs have a greater intracellular osmolaritythan the external medium because Hb as well as ionscontribute to the osmotic pressure of ion-permeableRBCs at ionic equilibrium. This osmotic imbalance pro-vides the driving force for water influx, creating increasesin mean corpuscular volume and, eventually, hemolysiswhen a critical RBC volume is reached. One mechanismby which Erythrosol protects phototreated RBCs may beby balancing osmotic differences inside and outsideRBCs through the presence of the impermeable citrateanion. Freedman and Hoffman found that adding 30 to35 mM of the impermeable oligosaccharide, sucrose, in-hibited hemolysis from ion-permeable RBCs.15 Erythro-sol contains a comparable concentration of citrate (26.6mM) and may function similarly. The finding that theMCV of DMMB-treated RBCs suspended in Adsol in-creased during storage concurrent with hemolysis whilesimilarly treated RBCs suspended in Erythrosol did notsupports the notion that cells in Adsol predominantlylyse via a colloidal osmotic mechanism. Finally, the re-sidual hemolysis in phototreated RBCs suspended in Ery-

throsol may be caused by an osmotic imbalance due to asuboptimal citrate concentration or, alternatively, mayoccur by some other mechanism.

Despite the development of a rapid, reproducible,and robust DMMB photoinactivation system, RBC stor-age properties were not preserved to permit 42-day stor-age. Improvement of RBC storage properties of pho-totreated cells may require optimized storage solutionsand/or the addition of RBC-specific antioxidants. TheRBC band 3 ligand dipyridamole16 has been shown to actas an antioxidant17-20 and protect RBCs from DMMB-induced photohemolysis without compromising extracel-lular virus kill.21 Perhaps these or other strategies to limitRBC membrane photodamage will improve storage prop-erties of DMMB-phototreated RBCs.

ACKNOWLEDGMENTS

We are appreciative of the efforts of Mary Stewart-Wesson and

Chong-Son Sun, PhD, Baxter Healthcare, for their assistance in

the initial characterization of the oxygen overlay and flow-cell

system.

REFERENCES

1. Schreiber GB, Busch MP, Kleinman SH, et al. The risk of

transfusion-transmitted viral infections. The Retrovirus

Epidemiology Donor Study. N Engl J Med 1996;334:1685-

90.

2. NHLBI. In vitro inactivation of viruses in blood compo-

nents. RFA HL-00-010. Bethesda: NHLBI; January 18,

2000.

3. Dodd RY, Notari EP, Stramer SL. Current prevalence and

incidence of infectious disease markers and estimated

window period risk in the American Red Cross blood do-

nor population. Transfusion 2002;42:975-9.

4. Horowitz B, Ben-Hur E. Efforts in minimizing risk of viral

transmission through viral inactivation. Ann Med 2000;32:

475-84.

5. Corash L. Inactivation of viruses, bacteria, protozoa and

leukocytes in platelet and red cell concentrates. Vox Sang

2000;78(Suppl 2):205-10.

6. Wagner SJ, Skripchenko A, Robinette D, Mallory DA, Cin-

cotta L. Preservation of red cell properties after virucidal

phototreatment with dimethylmethylene blue [published

erratum appears in Transfusion 1998;38:996]. Transfusion

1998;38:729-37.

7. Wagner SJ, Skripchenko A, Pugh JC, Suchmann DB, Ijaz

MK. Duck hepatitis B photoinactivation by dimethylmeth-

ylene blue in RBC suspensions. Transfusion 2001;41:1154-8.

8. Skripchenko A, Wagner SJ. Inactivation of WBCs in RBC

suspensions by photoactive phenothiazine dyes: compari-

son of dimethylmethylene blue and MB. Transfusion

2000;40:968-75.

9. Hogman CF, Eriksson L, Gong J, et al. Half-strength ci-

WAGNER ET AL.


trate CPD combined with a new additive solution for im-

proved storage of red blood cells suitable for clinical use.

Vox Sang 1993;65:271-8.

10. Hogman CF, Eriksson L, Wallvik J, Payrat JM. Clinical

and laboratory experience with erythrocyte and platelet

preparations from a 0.5CPD Erythro-Sol opti system. Vox

Sang 1997;73:212-9.

11. Standefer JC, Vanderjagt D. Use of tetramethylbenzidine

in plasma hemoglobin assay. Clin Chem 1977;23:749-51.

12. Meryman HT, Hornblower M, Keegan T, et al. Refriger-

ated storage of washed red cells. Vox Sang 1991;60:88-98.

13. Meryman HT, Hornblower M. Manipulating red cell intra-

and extracellular pH by washing. Vox Sang 1991;60:99-

104.

14. Pooler JP. The kinetics of colloid osmotic hemolysis. II.

photohemolysis. Biochim Biophys Acta 1985;812:199-205.

15. Freedman JC, Hoffman JF. Ionic and osmotic equilibria of

human red blood cells treated with nystatin. J Gen Phys-

iol 1979;74:157-85.

16. Grinstein S, Ship S, Rothstein A. Anion transport in rela-

tion to proteolytic dissection of band 3 protein. Biochim

Biophys Acta 1978;507:294-304.

17. Kusmic C, Picano E, Busceti CL, Petersen C, Barsacchi R.

The antioxidant drug dipyridamole spares the vitamin E

and thiols in red blood cells after oxidative stress. Cardio-

vasc Res 2000;47:510-4.

18. Ruggiero AC, Nepomuceno MF, Jacob RF, Dorta DJ,

Tabak M. Antioxidant effect of dipyridamole (DIP) and its

derivative RA 25 upon lipid peroxidation and hemolysis in

red blood cells. Physiol Chem Phys Med NMR 2000;32:35-

48.

19. Iuliano L, Pedersen JZ, Rotilio G, Violi F. A potent chain-

breaking antioxidant activity of the cardiovascular drug

dipyridamole. Free Rad Biol Med 1995;18:239-47.

20. Iuliano L, Ghiselli A, Alessandri C, Bonavita MS, Violi F.

Superoxide anion scavenging properties of dipyridamole.

Thromb Haemost 1989;61:149.

21. van Steveninck J, Trannoy LL, Besselink GA, et al. Selective

protection of RBCs against photodynamic damage by the

band 3 ligand dipyridamole. Transfusion 2000;40:1330-6.



use of a flow-cell system to investigate virucidal dimethylmethylene blue phototreatment in two...

Documents