Photochemistry und Phurobiologv, Vol. 62, No. 3, pp. 575-579. 1995 Printed in the United States. All rights reserved

003 1-8655195 $05.00+0.00 0 1995 American Society for Photobiology

BIODISTRIBUTION AND VIRUS INACTIVATION EFFICACY OF A SILICON PHTHALOCYANINE IN RED BLOOD CELL CONCENTRATES AS A

FUNCTION OF DELIVERY VEHICLE

E. BEN-HuR*I, M. M. ZUK], S. CHIN’, D. BANERJEE~, M. E. KENNEY~ and B. HOROWITZ’

‘New York Blood Center, 310 E. 67th Street, New York, NY 10021, USA and *Department of Chemistry, Case Western Reserve University, Cleveland, OH 44106, USA

(Received 21 March 1995; accepted 30 May 1995)

Abstract-The silicon phthalocyanine, HOSiPcOSi(CH,),(CH,),N(CH,), (Pc 4), is a new photosensi- tizer that can inactivate lipid-enveloped viruses in red blood cell concentrates (RBCC) upon exposure to red light. Because Pc 4 is insoluble in water, it was delivered either as an emulsion in saline and cremophor EL (CRM) or as a solution in dimethyl sulfoxide (DMSO). In RBCC, Pc 4 added in either vehicle distributed between the plasma and red blood cells (RBC) in a ratio of 4:6, similar to the ratio of these components in RBCC 3:7 (i.e. a hematocrit of 70%). Light exposure did not affect this distri- bution and caused only marginal degradation of Pc 4 at a light dose that inactivates >5 log,, vesicular stomatitis virus (VSV). Among human plasma proteins, Pc 4 bound mainly (about 70%) to lipoproteins and to a lesser extent to albumin and lower molecular weight proteins when delivered in DMSO. When delivered in CRM, distribution between lipoproteins and albumin became more even. Among the lipo- proteins Pc 4 bound almost exclusively to very low-density lipoproteins (VLDL) when delivered in DMSO and to both VLDL and low-density lipoproteins when added in CRM. The rate of VSV inacti- vation was independent of the delivery vehicle but there was less RBC damage, as measured by he- molysis during storage, when Pc 4 was added in CRM. These results indicate that using CRM as emulsifier can enhance the specificity of Pc 4-induced photochemical decontamination of RBCC for transfusion.

INTRODUCTION

Blood transfusion, although highly safe, does involve certain risks. One of these, the transmission of human pathogenic viruses such as human immunodeficiency virus (HIV),? is of particular public concern. Consequently, a major effort directed toward the development of blood sterilization tech- niques is underway. Currently, all blood products are avail- able in virally inactivated forms with the exception of red cell and platelet concentrates. The use of photosensitizers has shown promise for the inactivation of viruses in cellular blood components (see North et al.’ and Ben-Hur and Ho- rowitz2 for reviews). For red blood cell concentrates (RBCC), we have been using this approach with various phthalo~yanines.~-’ Members of this class of photosensitizer are promising because of their intense absorption band in the far red (650-700 nm). Among the phthalocyanines, HOS~PCOS~(CH,)~(CH~),N(CH,), (Pc 4) is of particular in- terest because in addition to having a high virucidal activity it also inactivates the parasites Trypanosoma cruzi and Plas-

*To whom correspondence should be addressed. tAbbreviations: CRM, cremophor EL; DMSO, dimethyl sulfoxide;

GSH, glutathione; HDL, high-density lipoprotein; HIV, human immunodeficiency virus; LDL, low-density lipoprotein; LPDP, li- poprotein-depleted plasma; PBS, phosphate-buffered saline; Pc 4, the silicon phthalocyanine HOSiPcOSi(CH,),(CH,)3N(CH3)2; PDT, photodynamic therapy; PLL, poly-L-lysine; RBC, red blood cell; RBCC, red blood cell concentrates; VLDL, very low-density lipoprotein; VSV, vesicular stomatitis virus.

modium falciparum,2 the causative agents of Chagas disease and malaria, respectively.

The compound Pc 4 is hydrophobic and requires solubi- lization in organic solvents such as dimethyl sulfoxide (DMSO). In this study, we report on the formulation of Pc 4 in an aqueous solution of cremophor EL (CRM, polyox- yethylene glycol triricinoleate) and compare its biodistribu- tion in RBCC with that of Pc 4 in DMSO. In addition, virus inactivation and red blood cell (RBC) damage upon light exposure were also compared when the Pc 4 is delivered by the two methods. Cremophor EL has been used for solubi- lization of hydrophobic drugs such as cyclosporin,8 t a x ~ l ~ . ’ ~ and photosensitizers.”’2 In the latter, CRM caused a shift in binding of a ketochlorin from albumin to low-density lipo- protein (LDL) and enhanced the photodynamic therapy (PDT) efficacy.I2 In the present work, we report on the bio- distribution of Pc 4 in RBCC. The compound Pc 4 was bound in human plasma to albumin and very low-density lipoprotein (VLDL) and this distribution was affected by CRM. Vesicular stomatitis virus (VSV) inactivation rate by Pc 4 photosensitization in RBCC was not altered by CRM. However, RBC damage, as assayed by delayed hemolysis of human RBC or circulatory survival of rabbit RBC, was less with CRM delivery of Pc 4.

MATERIALS AND METHODS

Chemicals. The compound Pc 4 was synthesized as previously described” and was purified by reversed-phase chromatography us-

576 E. BEN-HUR et a1

ing a CIS column. Its purity was about 97%. All other chemicals were analytical or HPLC grade, as required. Cremophor EL was purchased from Sigma Chemical Co. (St. Louis, MO). Stock solu- tions of Pc 4 at either 1 mM in DMSO or at 0.4 mM in 10% CRM in saline (see below) were made up and stored at -20°C.

Phototreatment of RBCC. Red blood cell concentrates were ob- tained from the New York Blood Center at 70% hematocrit. Samples of 3 mL RBCC were treated in polystyrene tubes. Irradiation was done at 25°C with red light using a xenon short arc lamp (Oriel Corp., Stratford, CT) equipped with a cut-off filter transmitting at A > 600 nm. Since no infrared filter was used the samples were cooled by an air stream. Sample temperature did not exceed 26°C. Light exposure was initiated 30 min after addition of Pc 4. When used, the quenchers glutathione (GSH), mannitol and Trolox (Sigma Chemical Co.) were at concentrations of 4, 4 and 5 mM, respec- tively. The irradiance at the samples was 25 mW/cm*, as measured with a photometer (model IL 1350, International Light, Newbury- port, MA). About 10% of the light energy was overlapping the ab- sorption spectrum of Pc 4. The RBCC were irradiated on a hema- tology mixer (Fisher Scientific, Inc., Pittsburgh, PA) to obtain even exposure.

VSV inacrivation. Inactivation of VSV in RBCC was studied us- ing a standard virus infectivity assay? Prior to photodynamic treat- ment, RBCC samples were spiked with lo7 infectious units per mL together with Pc 4 at 2 pM final concentration and quenchers at the concentrations given above. After treatment, samples were diluted 10-fold with Dulbecco's modified Eagle's medium containing 5% fetal calf serum and centrifuged at 1800 rpm for 5 min to remove the RBC. The supernatants were sterile filtered using 0.22 pm filters (Millipore, Bedford, MA) and either stored at -80°C or assayed immediately for VSV infectivity.

Formulation of Pc 4 in CRM. Three milligrams of Pc 4 was dissolved in 2 mL of methanol and then 1 mL of CRM prewarmed to 45°C was added dropwise to the methanol solution with vigorous mixing. Mixing was continued overnight at room temperature (to evaporate off the methanol), and the concentrate was diluted with 9 mL of phosphate-buffered saline (PBS) to obtain a final concentra- tion of 10% CRM and 0.4 mM Pc 4. The resulting emulsion was filter sterilized using a 0.45 pm filter and stored at -20°C. The residual methanol concentration in the emulsion was 50 ppm, as determined by gas chromatography. Its Pc 4 concentration was de- termined both by spectrophotometry and by HPLC.

Determination of Pc 4 in RBCC using HPLC. Samples of 2 mL RBCC containing 2 FM Pc 4 and the standard mixture of quenchers were centrifuged at 2500 rpm for 15 min. Aliquots of 0.6 mL plasma and 0.8 mL RBC were taken for analysis. The RBC were washed five times with PBS prior to extraction. The plasma and washed RBC aliquots were extracted twice with 2 mL of acetonitrile by vortexing for 1 rnin and ultrasonication for 10 min. The extracts were centrifuged at 2500 rpm for 10 min, the acetonitrile layer was reduced to a volume of 0.5 mL at 45-50°C under a stream of nitro- gen and 100 pL of 5% hexadecyltrimethylammonium bromide (Ce- trimide, Sigma) was added (to improve the recovery of Pc 4 in the next step). Samples were vortexed, diluted with 3 mL H,O and trans- ferred to C,* Sep-Pak ( 1 g) cartridges (Waters, Milford, MA) pre- washed with methanol and then water. The loaded cartridges were washed with 12 mL of water followed by 2 mL of methanol. Then, the Pc 4 was eluted with 18 mL methanol and the eluant was evap- orated to dryness under a stream of nitrogen at 45-50°C. The residue was dissolved in 1 mL methanol and aliquots of 10-100 pL were injected into a C,* analytical HPLC column, 5 Fm, 25 cm X 4.6 mm (Rainin, Emeryville, CA). The amount of Pc 4 in the effluent was determined on-line by absorption and fluorescence spectrosco- py. The mobile phase consisted of 10% H,O and 1% acetic acid in methanol and was isocratically pumped at a flow rate of 1.5 mL/ min. The retention time of Pc 4 was about 4 min. The HPLC system (Thermo Separation Products, Fremont, CA) was equipped with a fluorescence detector (model FL 2000) set at A,, = 610 nm and A,, = 670 nm as well as an absorption detector (Spectra Focus) set at 668 nm in order to monitor Pc 4. The HPLC system consisted of autosampler (model AS 3000), a high-pressure pump (model P4000) and SN4000 module and was operated by version 2.5 PCl000 soft- ware in 0 s - 2 system, Quantitation of Pc 4 in the analyzed samples was done wine external standards, i.e. a calibration curve was ob-

tained by injecting known amounts of Pc 4. The detection limit of Pc 4 under these conditions was 2 ng/mL of plasma or RBC and its recovery was 6%70%.

The determination of Pc 4 in RBCC was done in five samples for each experiment. The results are shown as the mean % standard errors. Statistical analysis of the results was performed using the two-tailed Student t-test (small sample analysis). Differences were considered statistically significant at P 5 0.05.

Distribution of Pc 4 among plasma proteins. Samples of plasma proteins derived from RBCC were separated according to molecular weights on a Sephacryl S-300 HR (Pharmacia Biotech, Piscataway, NJ) gel-filtration column (2.5 X 50 cm). A sample of 10 mL of the plasma was loaded onto the column and eluted with PBS. The ab- sorbance of the effluent was monitored on-line at 278 nm with a UV detector (Uvicord S, LKB) attached to a strip-chart recorder. Fractions of 5 mL were collected and the amount of Pc 4 in each was estimated by fluorometry at A,, = 607 nm and A,, = 680 nm after appropriate dilution with PBS, using a spectrofluorometer (Shi- madzu model RF 1501). The procedure was calibrated with molec- ular weight standards, yielding a linear curve over the range 2 5 4 6 9 kDa.

Plasma lipoproteins were separated by density-gradient ultracen- trifugation, as previously described.14 In short, plasma was adjusted to d = 1.26 g/mL with solid KBr. Three milliliters of the adjusted plasma was overlaid with a discontinuous KBr gradient comprised of 3 mL of d = 1.21 g/mL KBr solution followed by 3 mL of d =

1.063 g/mL and 3 mL of d = 1.006 g/mL KBr solution. The gradient was centrifuged at 40000 rpm (274 000 g) in a Beckman SW41 rotor at 10°C for 24 h. The lipoproteins floated at d = 1.006 g/mL, at d = 1.063 g/mL and at d = 1.21 g/mL were collected from the top of the gradient as VLDL, LDL and high-density lipoprotein (HDL) fractions. The remainder of the gradient was collected as lipoprotein- depleted plasma (LPDP). The amounts of Pc 4 in the fractions were estimated by fluorometry, as described above.

RBC hemolysis during storage. Samples of RBCC were stored after treatment with Pc 4 at 5°C in Vacutainer glass tubes (Becton Dickinson, Rutherford, NJ). The extent of hemolysis was determined by comparing the hemoglobin in the supernatant to the total he- moglobin. The total hemoglobin was determined using the Drabkin reagent (Sigma Procedure #525, Sigma Chemical Co.). The absorp- tion at 540 nm was used to calculate the amount of hemoglobin released in the supernatant. It should be noted that these are not standard blood-banking storage conditions and as a result RBC he- molysis was accelerated.

Determination of RBC negative surface charge. Samples of RBCC treated with Pc 4 in the absence of quenchers were assayed for the reduction of RBC surface negative charge using poly-L-lysine (PLL) binding technique, described previously." The quenchers were omitted to facilitate the observation of RBC damage. In the presence of the quenchers much higher light doses were required to cause observable reduction of RBC surface-negative charge.15

RESULTS

Distr ibut ion of Pc 4 b e t w e e n RBC and plasma

The distribution of Pc 4 between plasma and RBC in Pc 4-treated RBCC is shown in Table 1 . The dye was present in the plasma and bound to RBC in a ratio of 4:6, close to the ratio of these components in RBCC (3:7). This distri- bution did not change significantly with incubation time or with delivery vehicle. Light exposure of Pc 4-treated RBCC caused only minor (-15%) photodegradation of Pc 4 in both the plasma and RBC phases (Table 1) .

Binding of Pc 4 to plasma prote ins

The vehicle used to deliver Pc 4 to the RBCC had some effect on the binding of the photosensitizer to plasma protein fractions. Figure 1 shows the distribution of Pc 4 among plasma proteins in Pc 4-treated RBCC, as determined by gel filtration. Without light exposure, when added in DMSO,

Virus inactivation in red blood cells 577

Table 1. Distribution of Pc 4 between plasma and RBC*

Plasma RBC

CRM DMSO CRM DMSO Incubation time (min) Conc.? Amount$ %§ Conc. Amount % Conc. Amount 90 Conc. Amount %

0 1.36 2 0.11 0.41 f 0.03 41 1.40 5 0.08 0.42 2 0.02 46 0.83 2 0.06 0.58 f 0.04 59 0.70 2 0.09 0.49 2 0.06 54 5 1.38 ? 0.07 0.42 2 0.02 42 1.34 t 0.10 0.40 f 0.03 46 0.82 t 0.05 0.57 2 0.04 58 0.68 t 0.06 0.48 2 0.05 54 30 1.43 ? 0.05 0.43 f 0.02 42 1.20 f 0.03 0.36 t 0.01 44 0.84 f 0.04 0.59 2 0.02 58 0.66 f 0.06 0.47 ? 0.04 56 90 1.12 2 0.06 0.34 2 0.02 36 1.31 2 0.04 0.39 2 0.01 46 0.84 ? 0.04 0.59 2 0.03 64 0.67 ? 0.04 0.47 2 0.03 54 30 + 60u

(90 J/cm*) 1.00 2 0.08 0.30 2 0.04 37 1.03 2 0.04 0.31jI t 0.01 43 0.72 2 0.05 0.51# ? 0.04 63 0.59 2 0.04 0.42% ? 0.03 57

*The levels of Pc 4 were determined with HPLC as described in the Materials and Methods. Total amount of Pc 4 added was 1.434 pg/mL

TThe concentration of Pc 4 (kg/mL) in the blood component analyzed. +Amount of Pc 4 (pg) in blood component analyzed in 1 mL RBCC taking into account that plasma comprises 30% of RBCC and RBC is

§Percentage of Pc 4 recovered in the blood component of the total recovered in 1 mL RBCC. JIStatistically significant (using two-tailed Student’s t test) difference comparing 90 min dark- to light-treated samples at P 5 0.001. #P 5 0.01. s[P 5 0.05.

and recovery was 6&70%.

70%.

about two thirds of the dye was bound to high molecular weight proteins (MW >440 kDa), presumably lipoproteins. The rest was bound to albumin and lower molecular weight proteins. Delivery of Pc 4 in CRM caused a shift of the early-eluting Pc 4 to a higher MW peak and the fraction of dye in this peak was reduced to about half the total. With light exposure, the amount of Pc 4 fluorescence associated with the high MW peak was reduced by about a third, both when delivered in DMSO (Fig. 1) and in CRM (not shown for clarity).

Further analysis of Pc 4 binding to lipoproteins was per-

0.8

0.6

0.4

0.2

0 5 10 15 20

Fraction Number Figure 1. Distribution of Pc 4 among plasma proteins separated by gel filtration. The RBCC at 70% hematocrit were treated for 30 min with the standard quenchers mixture and 2 Pc 4 in DMSO (cir- cles) or CRM (triangles) in the dark (filled symbols) or with 90 .I/ cm2 red light (open symbols). Plasma (10 mL) was then applied to a Sephacryl S-300 column and proteins were separated as described in the Materials and Methods. The full line is the tracing of protein absorption and dashed lines are Pc 4 fluorescence at 680 nm. The former was essentiallv identical for all three experiments shown.

formed with the aid of density gradient separation. The re- sults (Table 2) show that when delivered in DMSO Pc 4 is bound almost exclusively to VLDL among the lipoproteins. Administration of the dye in CRM resulted in some binding to LDL as well.

VSV inactivation and RBC damage by Pc 4

The ability of Pc 4 to photoinactivate cell-free VSV in RBCC when delivered in DMSO or CRM was compared (Fig. 2). Evidently, the delivery vehicle had no effect on the rate of VSV kill. There was no difference in the potential of Pc 4, delivered in either DMSO or CRM, to induce RBC damage when reduction of negative surface charge was as- sayed (Fig. 3). However, hemolysis of RBC was reduced during storage when Pc 4 was delivered in CRM (Fig. 4). In addition, survival of rabbit RBC increased significantly when delivery of Pc 4 was in CRM rather than in DMSO (Table 3).

DISCUSSION

The compound Pc 4 is one of the most effective phthal- ocyanine photosensitizers for the inactivation of viruses and

Table 2. Distribution of Pc 4 among plasma lipoproteins*

Pc 4 (% of total)

Density (g/mL) Designation DMSO CRM ~~

5 1.006 VLDL 63.8 2 3.9 39.8 2 3.8 1.00&1.063 LDL 1.1 2 0.2 11.1 2 0.9 1.063-1.2 1 HDL <0.2 0.9 2 0.1

21.21 LPDP 35.1 2 2.3 48.2 2 3.1

*The plasma was derived from RBCC treated for 30 min with the standard quenchers mixture and 2 pA4 Pc 4 delivered in DMSO or CRM. The amounts of Pc 4 in the fractions were determined by fluorometry as described in the Materials and Methods. En- tries are percentage of the total Pc 4 detected in the gradients and SEM. Recovery of Pc 4 in the gradients was -90%. Results are the mean ? SEM of separate experiments (n = 3) with nc light exposure.

E. BEN-HIJR er a[.

0 1 0 3 0 5 0 7 0 9 0

L i g h t Fluence (Jlcm’)

Figure 2. Photosensitized inactivation of VSV. The RBCC contain- ing Trolox, GSH and mannitol as quenchers and 2 p M Pc 4 added either in DMSO (open circles) or CRM (filled circles) were spiked with VSV, incubated for 30 min and exposed to red light at the indicated fluence. Results are the mean of separate experiments (n = 3 ) 2 SEM.

parasites in RBCC.’r ’6 Cell-free, cell-associated and latent HIV are inactivated at an even lower light fluence than VSV when Pc 4 is used.” In previous work we have administered Pc 4 from a DMSO stock solution. However, for blood- banking purposes, administration must be made through a sterile docking device with an aqueous medium. We have therefore formulated Pc 4 in 10% CRM as a delivery vehicle and compared it with delivery in DMSO. The results dem-

I

0 3 0 6 0 9 0 1 2 0 1 5 0

L i g h t Fluence ( J lcm’)

Figure 3. Photosensitized reduction of RBC negative surface charge. The RBCC diluted two-fold with PBS were irradiated in the absence of quenchers with 2 )uM Pc 4 added either in DMSO (open circles) or CRM (filled circles). Negative surface charge was quantitated by binding of RBC to PLL, as previously described.“ These less than optimal treatment conditions were used to facilitate the observation of damage. Results are of a single experiment. About 90% of the untreated cells bind to PLL.

0 a Control

--O--- Pc4 inCRM

U Pc4inDMSO

2 V M PC4 + 90 J/Cm2

I

0 5 1 0 1 5 2 0 2 5

S t o r a g e Time (days)

Figure 4. Hemolysis of photosensitized RBC during storage. The RBCC containing Trolox, GSH and mannitol as quenchers and 2 p M Pc 4 added either in DMSO or CRM, or not treated, as indicated. were exposed to 90 J/cm2 red light. Hemolysis was followed during storage of RBC as described in the Materials and Methods. Results are the mean of separate experiments (n = 3) 2 SEM.

onstrate that delivery in CRM yields the same VSV inacti- vation (Fig. 2) but causes less RBC damage as reflected by reduced hemolysis during storage (Fig. 4) and increased RBC survivability in vivo (Table 3). The latter result was somewhat unexpected because the fraction of Pc 4 bound to RBC when delivered in CRM is slightly higher than when DMSO is used (Table 1). It is possible that when delivered in CRM, Pc 4 either binds to sites on RBC that are less susceptible to photodynamic damage and/or it is more ag- gregated upon binding (aggregated phthalocyanines are pho- tochemically inactive).

Delivery of Pc 4 in CRM instead of DMSO causes some changes in the distribution of the Pc 4 among the plasma proteins. Thus, with CRM delivery more Pc 4 is bound to albumin and lower MW proteins (Fig. 1). In addition, there is a shift in the binding of Pc 4 from VLDL to LDL and HDL, although most of it is still bound to the former (Table 2). Other studies on the distribution of photosensitizera among plasma proteins as a function of delivery vehicle also show that the distribution depends on the sensitizer. Thus. while CRM has only minor effects on the distribution of a mesoporphyrin among the albumin and lipoprotein fractions

Table 3. Survival of rabbit RBC in vivo after Pc 4 PDT*

None 2 p M Pc 4, no light 2 p M Pc 4 in DMSO + 90 J/cm* 2 p M Pc 4 in CRM + 90 J/cm2

8.2 8.0 2.5 3.8

~ ~ _ _ _ _ ~

*Rabbit RBCC were treated, and circulatory survival was deter- mined as described previously! Results are of a single experi- ment.

+TS0 is the circulatory half-life of RBC.

Virus inactivation in red blood cells 579

of human plasma,'" a ketochlorin, a substantially more hy- drophobic sensitizer, is shifted in its binding from albumin to HDL at low CRM concentrations and to LDL at higher concentrations.Iv The concentration of CRM used in the present study (0.05%) is intermediate compared to those used by earlier workers (0.008-0.16%). Previously, it was also noted that at high concentrations (>0.09%) CRM de- grades plasma lipoproteins, resulting in the appearance of new species that can mimic the behavior of LDL in a den- sity-gradient ~ y s t e m . ~ * , ~ ~ , ~ ~ To what extent this happened un- der our treatment conditions remains to be seen. Because CRM is localized in the LDL fraction> in a density gradientI9 this may contribute to the shift observed (Table 2) . When ZnPc is delivered in liposomes it is incorporated primarily into HDL and LDL and only to a small extent into VLDL? For a recent review of the distribution of photosensitizers among human plasma proteins see Kongshaug.2'

An important point that emerges from this study is the observation that there is little photodegradation of Pc 4 under conditions leading to >5 log,,, kill of VSV (Table 1). The reduced fluorescence of Pc 4 associated with lipoproteins after light exposure (Fig. 1) may therefore reflect changes in the microenvironment of the dye, which lead to a reduced fluorescence quantum yield. Data derived from the density- gradient centrifugation studies show no change in the distri- bution of Pc 4 after light exposure, but again, the total flu- orescence associated with VLDL is reduced by about a third (data not shown). From the toxicological point of view, the minimal extent of photodegradation is favorable, as there is no need to contend with the potential toxicity of the photo- products.

In conclusion, Pc 4 delivered in CRM is as effective for virus sterilization of RBCC as when delivered in DMSO. At the same time, RBC damage is reduced with CRM. Thus, the use of CRM as a vehicle, in addition to making Pc 4 dispersible in water, has the advantage of enhancing the specificity of the treatment.

Acknowledgements-The expert technical assistance of S. Kirschen- baum and J. Oetjen is gratefully acknowledged. This work was sup- ported in part by award no. 2R01-HL41221 from the National Heart, Lung and Blood Institute, and award no. Pol-CA48735 from the National Cancer Institute.

REFERENCES

1. North, J., H. Neyndorff and I. G. Levy (1993) Photosensitizers as virucidal agents. J. Phorochem. Photobiol. B Biol. 17, 99- 108.

2. Ben-Hur, E. and B. Horowitz (1995) Advances in photochem- ical approaches for blood sterilization. Phorochem. Photobiol. 62, 383-388.

3. Horowitz, B., B. Williams, S. Rywkin, A. M. Prince, D. Pascual, N. Geacintov and J. Valinsky (1991) Inactivation of viruses in blood with aluminum phthalocyanine derivatives. Transfusion 31, 102-108.

4. Rywkin, S., L. Lenny, J. Goldstein, N. Geacintov, H. Margolis-

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16

17

Nunno and B. Horowitz (1992) Importance of type I and type I1 mechanisms in the photodynamic inactivation in blood with aluminum phthalocyanine derivatives. Phorochem. Photobiol. 56, 463469. Ben-Hur, E., R. C. Hoeben, H. Van Orrnondt, T. M. A. R. Dubbelman and J. Van Steveninck (1992) Photodynamic inac- tivation of retroviruses by phthalocyanines: the effect of sulfo- nation, metal ligand and fluoride. J. Photochem. Phorobiol. B Biol. 13, 145-152. Rywkin, S., E. Ben-Hur, Z . Malik, A. M. Prince, Y. S. Li, M. E. Kenney, N. L. Oleinick and B. Horowitz (1994) New phthal- ocyanines for photodynamic virus inactivation in red blood cell concentrates. Photochem. Phorobiol. 60, 165-1 70. Smetana, Z . , E. Mendelson, J. Manor, J. E. van Lier, E. Ben- Hur, S. Salzberg and Z . Malik (1994) Photodynamic inactiva- tion of herpes viruses with phthalocyanine derivatives. J. Pho- tochem. Phorobiol. B Biol. 22, 37-43. Cohen, D. J., R. Loertscher, M. F. Rubin, N. L. Tilney, C. B. Carpenter and T . B. Strom (1984) Cyclosporin: a new immu- nosuppressive agent for organ transplantation. Ann. Intern. Med. 101, 667-682. Slichenmyer, W. J. and D. D. von Hoff (1991) Taxol: a new and effective anti-cancer drug. Anti-Cancer Drugs 2, 5 19-530. Wiernik, P. H., E. L. Schwartz, J. J. Strauman, J. P. Dutcher, R. B. Lipton and E. Paietta (1987) Phase I clinical and phar- macokinetic study of taxol. Cancer Res. 47, 2486-2493. Morgan, A,, G. M. Garbo, R. W. Keck and S. H. Selman (1988) New photosensitizers for photodynamic therapy: combined ef- fect of metallopurpurin derivatives and light on transplantable bladder tumors. Cancer Res. 48, 194-1923, Woodburn, K., C. K. Chang, S. Lee, B. Henderson and D. Kes- sel ( 1 994) Biodistribution and PDT efficacy of a ketochlorin photosensitizer as a function of the delivery vehicle. Phorochem. Photobiol. 60, 154-159. Oleinick, N. L., A. R. Antunez, M. E. Clay, B. D. Rihter and M. E. Kenney (1993) New phthalocyanine photosensitizers for photodynamic therapy. Photochem. Photobiol. 57, 242-247. Redgrave, T . G., D. C. K. Roberts and C . E. West (1975) Sep- aration of plasma lipoproteins by density-gradient ultracentri- fugation. Analyt. Biochem. 65, 42-49. Ben-Hur, E., S. Rywkin, I. Rosenthal, N. E. Geacintov and B. Horowitz (1995) Virus inactivation in red blood cell concen- trates by photosensitization with phthalocyanines: protection of red cells but not of vesicular stomatitis virus with a water sol- uble analog of vitamin E. Transfusion 35, 401406. Gottlieb, P., E. Ben-Hur, S. Lustigman, H. Margolis-Nunno and B. Horowitz (1994) Parasite inactivation in red blood cells and platelet concentrates by photodynamic treatment. Trunsfusion 34, 43s. [Abstract] Ben-Hur, E., A. C. E. Moor, H. Margolis-Nunno, P. Gottlieb, M. M. Zuk, S . Lustigman, B. Horowitz, A. Brand, J. Van Stev- eninck and T . M. A. R. Dubbelman (1996) Photosensitization- mechanisms and uses in transfusion medicine of cellular com- oonents. Transfusion Med. Rev. (In press)

18. Woodburn, K." and D. Kessel (1994) The alteration of plasma lipoproteins by cremophor EL. J. Phorochem. Photobiol. B Biol. 22, 197-201.

19. Kongshaug, M., L. S. Cheng, J. Moan and C . Rimington (1991) Interaction of cremophor EL with human plasma. Znt. J . Bio- chem. 23, 473-478.

20. Rensen, P. C. N., W. G. Love and P. W. Taylor (1994) In vitro interaction of zinc (11)-phthalocyanine-containing liposomes and plasma proteins. J. Photochem. Photobiol. B Biol. 26, 29-35.

21. Kongshaug, M. (1992) Distribution of tetrapyrrole photosensi- tizers among human plasma proteins. Int. J. Biochem. 24, 1239- 1265.