Effective and less toxic reversal of low-molecular weight heparin anticoagulation by a designer variant of protamine

Effective and less toxic reversal of low-molecular weight heparin anticoagulation by a designer variant of protamine T h o m a s W. Wakefield, M D , Philip C. Andrews, P h D , Shirley K. Wrobleski, BS, A m y M. Kadell, BS, Regina Schmidt, BS, Samir Tejwani, and James C. Stanley, M D , Ann Arbor, Mich.

Purpose: This investigation assessed protamine reversal of heparin anticoagulation by formation of a protamine-heparin alpha-helix by use of a new designer-variant protamine [ + 18BE] that was made from an existing protamine variant [ + 18B] whose non-alphahelix-forming amino acid proline (P) was replaced by an alpha-helix-forming glutamic acid (E). The rate of administration of the new [ + 18BE] variant protamine on efficacy and toxicity in comparison to that of [ + 21] standard protamine and [ + 18B] was also studied. Methods: AcetyI-EAA(K2A2K2A)4K2-Amide [ + 18BE] was administered intravenously in a 1 : 1 dose to low-molecular-weight heparin (LMWH)-anticoagulated (intravenous 150 IU antifactor Xa/kg) dogs over 10 seconds or 3 minutes (n = 7, each group). Reversal efficacy was documented by measuring activated clotting time, thrombin clotting time, antifactor Xa, and antifactor IIa. Toxicity was defined by measuring systemic blood pressure, heart rate, cardiac output, pulmonary artery pressure, and oxygen consumption. Measurements were made at baseline, after administration of LMWH, before its reversal, and for 30 minutes thereafter. Results were compared with those after L M W H reversal with [ + 21] standard protamine and the [ + 18B] variant. A total toxicity score (TTS) was calculated for each compound from maximal declines in blood pressure, heart rate, cardiac output, and oxygen consumption. Results: LMWH anticoagulation reversal was significantly (p < 0.01) less toxic over 10 seconds and 3 minutes with the [+ 18BE] designer variant (TI'S - 2.3, - 2.2) compared with the [+21] standard protamine (TTS - 6 . 4 , -7.2). Percent L M W H reversal at 3 minutes revealed [ + 18BE] to have antifactor Xa activity as high as 91%, compared with 68% for protamine [+21], when given over 3 minutes (p < 0.05). Conclusions: This investigation documents that a new designer variant of protamine [ + 18BE] has superior efficacy compared with [ + 21] standard protamine for reversal of L M W H anticoagulation and that this occurs with a highly favorable toxicity profile. (J VASC SURG 1995;21:839-50.)

Protamine reversal o f heparin or low-molecularweight heparin ( L M W H ) anticoagulation may be associated with adverse hemodynamic and hematologic events both experimentally in animal models and clinically in patients. >4 Although these effects are

From The Jobst VascularLaboratory,Sectionof VascnlarSurgery, Department of Surgery, and the Department of Biochemistry (Dr. Andrews), University of Michigan and the Ann Arbor Veterans Administration Medical Center, Ann Arbor. Supported by a Grant from the Veterans Administration, Merit No. 331. Presented at the Eighteenth Annual Meeting of the Midwestern Vascular SurgicalSociety,Sept. 23-24, 1994, Cincinnati, Ohio. Reprint requests: Thomas W. Wakefield,MD, UniversityHospital, 2210 THCC, 1500 E. MedicalCenter Dr., Ann Arbor, MI 48109-0329. 24/6/62360

often only temporary, severe nonimmunologic reactions, including death, have been associated with protamine administration. In a prior study, we found that a protamine variant with a + 18 charge because oflysine residues with acetyl and amide groups on its termini [ + 1 8 B ] had equal efficacy in L M W H reversal to [ +21] standard protamine. 4 Although a reduced toxicity accompanied use o f the [ + 18B] compound, it was still hazardous. It is known that when peptides bind to heparin, they form alpha (~) helix configurations, s,6 It was hypothesized that protamine peptide variants that would be more prone to c~ helix formation might provide better and less toxic L M W H reversal. This investigation was undertaken to determine whether the substitution o f the non-~ helix-forming amino 839

840

JOURNAL OF VASCULAR SURGERY May 1995

Wakefield et al.

Table I. Maximal mean declines in hemodynamic parameters Group

M A P (mm Hg) *

Protamine [ + 21] (10 seconds) [ + 18B] (10 seconds) [ + 18BE] (10 seconds) Protamine [ + 21] (3 minutes) [ + 18BE] (3 minutes)

- 32 - 10 - 8 - 18 0

CO (%) t -

VO 2 (%) {

32 18 11 28 -7

-

26 12 12 23 10

H R (beats/rain) - 18 -9 - 13 - 36 -5

up < 0.01 A N O V A between all groups. tp < 0.05 A N O V A between all groups.

Table II. Percent reversal of coagulation studies and platelet count changes Group

ACT

Antifactor Xa

TCT

Antifactor lIa

Platelet count

3 minutes Protamine [ + 21] (10 seconds) [ + 18B] (10 seconds) [ + 18BE] (10 seconds) Protamine [ + 2 1 ] (3 minutes) [ + 18BE] (3 minutes)

* 96 87 65 99 89

* 63 64 73 68 91

ns 99 99 100 100 100

ns 96 95 99 99 94

ns - 43 - 56 44 -32 -49

10 minutes Prbtamine [ + 21] (10 seconds) [ + 18B] (10 seconds) [ + 18BE] (10 seconds) Protamine [ + 2 1 ] (3 minutes) [ + 18BE] (3 minutes)

ns 83 87 78 87 95

* 45 34 51 53 68

* 98 95 95 99 100

* 72 83 98 99

ns - 36 - 34 - 3 - 10 - 8

30 minutes Protamine [ + 21] (10 seconds) [ + 18B] (10 seconds) [ + 18BE] (10 seconds) Protamine [ + 21] (3 minutes) [ + 18BE] (3 minutes)

* 78 98 81 80 97

ns 44 52 59 45 60

ns 96 100 95 98 99

ns 86 74 79 92 84

ns - 25 15 20 - 19 35

*p < 0.01 A N O V A between all groups.

acid proline (P) in [ + 18B] by the ci helix-forming amino acid glutamic acid (E), producing the designer-variant [ + 18BE], would improve efficacy even further while reducing toxicity for LMWH reversal. MATERIAL AND M E T H O D S Protamine-variant peptides were synthesized on preloaded Tentagel (Rapp Polymer, Tubingen, Germany) with 9-fluorenylmethoxycarbonyl-aminoacid derivatives in an Applied Biosystems 433 peptide synthesizer by use of conditional &protection plus single coupling (Applied Biosystems, Foster City, Calif.). The hydroxybenzotriazolyl esters of the 9-fluorenylmethoxycarbonyl,amino acids were formed with use of 2-(1H benzotriazol-l-yl)-l,l,3,3-tetramethyluronium hexafluorophosphate as an activation agent. Coupling and deprotection of the nascent peptide chains were performed under standard conditions for the synthesizer (FastMOC cycles) except as indicated above. Cleavage and final deprotection oc-

curred in 90% trifluoroacetic acid containing 5% ethanedithiol, 2.5 % thioanisole, and 2.5 % anisole for 2 hours at room temperature. Peptides were then precipitated from the trifluoroacetic acid by addition of 20 volumes of diethylether at - 2 0 ~ C. These peptides were purified by reversed-phase highperformance liquid chromatography (HPLC) on a 5 cm • 25 cm preparative reversed-phase column (Rainin, Woburn, Mass.) at a flow rate of 17 ml/min x 90 minutes with a gradient from 5% to 60% acetonitrile. The peptides were subsequently desalted on Sephadex (Pharmacia, Piscataway, N.J.) G-15 gel filtration columns (Pharmacia) equilibrated with 1N acetic acid and then freeze-dried. Each peptide was characterized by amino acid analysis, analytical reversed-phase HPLC, and mass spectroscopy to confirm purity and structure before use. Norleucine inclusion as an internal standard allowed accurate assessment of peptide concentration by amino acid analysis. Peptides were synthesized so that the number of

JOURNAL OF VASCULAR SURGERY Volume 21~ Number 5

lysine (K) residues determined the total peptide cationic charge, with standard salmine protamine's primary charge being [ + 21]. The protamine-variant peptides constructed for study included [ + 18BE] acetyl-EAA(KaA2K2A)4K2Amide and [ + 18B] Acetyl-PA(K2A2K2A)4K2amide. We have previously reported on [ + 18B], a + 18 compound in which the ends were acetylated and amidated to prevent in vivo degradation by aminopeptidases and carboxypeptidases, as well as increase its dipole moment and cr helix formation on binding to heparin. 4 The replacement of proline (P) with glutamic acid (E) in [ + 18BE] was undertaken to increase cx helix formation on binding with heparin. Also, an additional alanine was added for spacing considerations. Protamine's [ + 21] molecular structure is PR4S3RPVRsPRVSR6G2R4. Lysine, like arginine (R) in protamine, carries a positive charge of [ + 1] at physiologic pH. It is similar in size to arginine, yet its use in peptide synthesis avoids technical difficulties associated with automated production of proteins containing large numbers of arginine residues. Amino acid and peptide charge assignment was based on the known pKa of 9.7 for lysine and 12.5 for arginine. At physiologic pH 7.4, more than 99% of the lysine and arginine residues carry a charge of [ + 1]. In these protamine variants, the alanine (A) connecting amino acids increased the propensity for helix formation on binding to LMWH. Each variant peptide, as well as standard protamine, was tested in seven adult female mongrel dogs. The sequence of the peptide testing was random. Animals (mean weight 13 kg) were anesthetized by use of 15 mg/kg of sodium pentobarbital, intubated, and mechanically ventilated. Supplemental oxygen 4 L/min was supplied. The animals were hydrated with an intravenous infusion of lactated Ringer's solution, 22 ml/kg/hr. All animals were housed and cared for in the University of Michigan Unit for Laboratory Animal Medicine under the direction of a veterinarian according to the guidelines of the "Principles of Laboratory Animal Care" (National Society for Medical Research) and Guide for the Care and Use of Laboratory Animals (NIH Publication No. 86-23, revised 1985). All animals were anticoagulated with intravenous LMWH (Logiparin, Novo, Denmark) 150 IU/kg antifactor Xa activity. The protamine variant peptides and standard protamine (Eli Lilly, Indianapolis, Ind.), 1.5 mg/kg (100 IU/mg) were individually administered 30 minutes later. These compounds were injected intravenously rapidly over 10 seconds

Wakefield et al.

841

-7-6-5-4-3-2-

-1 0

O ~1 ~" 4.

O

~oO

O

ro

CO

ca

~ +

ca

cO

co

n

Fig. 1. TTS for protamine (prot) [ + 21], [ + 18B] variant, and the [ + 18BE] variant, after 10-second and 3-minute administration. to maximize hemodynamic effects, as well as more slowly over 3 minutes. The effects of injected compounds [ + 18B] and protamine [ +21] administered rapidly over 10 seconds have been presented before,4 but are included in this study for comparison to rapid injection of [ + 18BE] and slower injections of protamine [ + 21] and [ + 18BE]. Hemodynamic observations included measurement of systemic mean arterial blood pressure (MAP) and heart rate (HR) by a carotid artery catheter; pulmonary artery systolic and diastolic pressures and mixed venous oxygen saturataon (SvO2) by an oximetric pulmonary artery catheter placed through a femoral vein (Abbott Laboratories, North Chicago, Ill.); and systemic arterial oxygen saturation (SaO2) by a catheter positioned in the femoral artery. Cardiac output (CO) was measured with an electromagnetic square wave flow probe (Narcomatic, Houston, Texas) by quantitating pulmonary artery flow. These measurements allowed calculation of systemic oxygen consumption (902) by Fick equation relationships, with oxygen consumption = Flow x Hemoglobin x 1.34 [SaO 2 - SvO2]. All hemodynamic data, both measured and calculated, were collected and assessed with an on-line computer program (Workbench; Strawberry Tree, Sunnyvale, Calif.). Measurements and calculations were made at baseline, before administration of LMWH, 3 minutes before reversal, during reversal, every 10 seconds

J O U R N A L OF V A S C U L A R S U R G E R Y

842

May 1995

Wakefield et al.

15

30

~0 -5

2'

0,01

"

9

m

2~1 o o o g oo

-30

o

.

.

.

.

.

.

.

. o

10-

10-

Ot

0{

o o

o

[ I, o o

,

, o

5-

~

gc O

"~

-5-

-5 -

'~

-10

'~

-20-

-15 -15

-20-25

o

-30-35

--B--

-25-

~ i

i

i

i

-30[+18BE](10sec)

-i

Prot[+21](10sec) [+18B](10sec)

-35;

o

i

g

Time (sec)

i

o

iI

g

i

o

i

~

i

g

(min)

-40

o g. . g . g. o. . g . g. ~ o o~o ~ oJ I , o o , g, T i m e (sec)

(min)

Fig. 2. MAP, CO, 902, and HR changes for protamine (prot) [+21]), [+ 18B] variant, and the [+ 18BE] variant, after 10-second administration. for 5 minutes after reversal, and at 10, 20, and 30 minutes after reversal. Although measurements were made every 10 seconds, only the values every 30 seconds were recorded and presented in this study, A total toxicity score (TTS) was calculated for each experiment, that was determined by the maximum declines in MAP, CO, VO2, and H R during the first 5 minutes after reversal. This time was chosen as the maximal hemodynamic effects occur during this interval. The maximum declines in the individual dog divided by the standard deviation derived from the animals in all groups combined were added to compute the TTS for each individual dog. In each group, the mean of the seven individual dogs was then calculated to obtain a TTS for each peptide studied. The more negative the TTS value, the more toxic the compound during LMWH reversal. Coagulation and hematologic studies included activated clotting time (ACT), heparin antifactor Xa activity, thrombin dotting time (TCT), heparin antifactor IIa activity, platelet count, and white blood cell count. These studies were performed in venous blood samples before LMWH administration, 3 minutes before reversal, and 3, 10, and 30 minutes after salmine and variant-protamine administration. ACT measurements were made with 2 ml of whole

blood collected in celite-containing tubes (Hemochron, International Technidyne, Edison, N.J.), the remainder of the blood was placed into standard citrated tubes and either stored on ice for heparin antifactor Xa and antifactor IIa determinations or centrifuged at 900 rpm for 10 minutes to obtain platelet-rich plasma (PRP). TCT studies used 0.2 ml PRY' and 0.1 ml TCT reagent (American Dade, Miami, Fla.) and were performed on a fibrometer (Baxter, Miami, Fla.) and run in duplicate with the values averaged. Platelet and white blood cell counts were analyzed by manual hemocytometry (Unopette; Becton-Dickenson, Rutherford, N.J.). Heparin antifactor Xa (FXa) and antifactor IIa (FIIa) assays were performed on citrated blood stored on ice after cold (4~ C) centrifugation at 2900 rpm for 20 minutes. This resulted in a platelet-poor plasma supernatant. In the FXa assay (Coatest Heparin; Kabi Vitrum, Stockholm, Sweden), excess antithrombin III followed by factor Xa was added to the platelet-poor plasma, which was then reacted with the chromogenic substrate S-2222. The FIIa assay was similarly performed with addition of thrombin (IIa) followed by the chromogenic substrate S-2238. Heparin antifactor Xa or antifactor IIa activity was inversely proportional to absorbance at

JOURNAL OF VASCULAR SURGERY Volume 21, Number 5

Wakefield et al.

15105-

30-

0~

10-

-5n

8

"~ -10g c

8

843

>

0!

8., ,g

-15-

-20-

,~ q0-

-25-

-20-

-30-

...... II . . . . . . . . . . o o o o o o 8 o o o o03o (D o oObo O,l o ou3o e0 ~ o~_.l

-35

.

.

.

.

~

II,

,

r

-3o

ea e4 o3

...... I I .......... II, , 9 oooooo~oooooooooooooo~.,..o~ ~

10-

10-

5-

5-

~ ~ ~ c,4 o4 t'N c0

0~

g 8 8, g 8

-5-

i

-5-10-15g -20-25 .c O -30-35-

-10-15-20-25-

-30 i -35

II . . . . . . . . . .

. . . . .

II . . . .

~

y

+

40

[+t 8BE](amin)

~ 1 7 6 1 7 6 1 7 6 1 7 6 1 7 6 1 7 6 1 7 6 1 7 6 1 7 6 1 7 o6 o1. o.7. o6. .o~o1. oc47 o6Loo1c0~7II03o6. .~ 1. .7o. .6o. .1S. .7o 6a 1~ 7 6 .

.

.

.

.

.

r

e4 oJ r

Time (so0)

During administration

~

.

.

(mini

After administration

.

.

II,

Time (sec)

During administration

Fig. 3. M A P , CO, HO2, and H I { changes for protamine (prot) [ + 2 1 ]

,

,

.

(mini

After administration

and the [ + 18BE]

variant after 3-minute administration.

405 nm, which indicated uninhibited factor Xa or IIa remaining in the sample. Circular dichroism spectra were run for [ + 18BE] on a Jasco (Easton, Md.) 710 CD instrument with a Peltier temperature controller. The spectra were scanned from 300 to 180 nm with four averaged acquisitions. One level of curve smoothing was applied to each averaged spectrum. The scan time was 50 nm/min; nitrogen flow in the sample compartment was 30 L/min; and the photomultiplier voltage was maintained below 600 V at wavelengths above 184 nm. The cell path length was 1 mm, and the sample was maintained at 25 ~ C. Samples were prepared just before spectra were acquired. All samples were prepared in HPLC-grade water or in trifluoroethanol (Aldrich Chemical Corp., Milwaukee, Wis.) in water. The final concentration of heparin and peptide was 0.1 mg/ml. Group data are expressed as mean _ SD. For all coagulation results, reversal over 100% was considered 100%, and was recalculated from previously published data for the groups protamine [ + 21] over 10 seconds and [ + 18B].4 Statistical analysis included analysis of variance (ANOVA) and unpaired two-

tailed Student's t tests where appropriate (Statworks; Cricket Software, Philadelphia, Pa.). RESULTS H e m o d y n a m i c changes. The most toxic compound for reversal of L M W H was standard protamine [+21] with a TTS 6.4 ___ 3.8 (10 seconds) and 7.2 _~ 3.5 (3 minutes) compared with - 3 . 1 _ 1.5 for [+18BI (I0 seconds), - 2 . 3 _+ 3.6 [+18BE] (10 seconds) and 2.2 _+ 1.7 for [+18BE] (3 minutes). These TTS scores were significantly different by ANOVA (p < 0.01) (Fig. 1). Specifically, [+18BE] (3 minutes) was different than protamine given over 10 seconds (p < 0.05); and [+18B] (10 seconds), [+18BE] (10 seconds) and [ + 18BE] (3 minutes) were different than protamine given over 3 minutes (p < 0.05, 0.05, and 0.01, respectively). Maximal mean declines in MAP, CO, VO2, and H R were defined ITable I, Figs. 2 and 3). None of the variants caused any increase in pulmonary artery pressure compared with protamine given over i0 seconds or 3 minutes (Fig. 4). Coagulation changes. Excellent efficacy of

844

JOURNALOF VASCULARSURGERY May 1995

Wakefield et al.

14-

14-

12

12-

10-

~10-

6

u)

4-

4o

2-

0t -2

ooo

-2

ooooo

14

~ . . . . o ~ o o o

41 . . . . . . . .

, ,11, , o o ~ o o o o

,

14-

12----II-- Pr0t[+21](10sec)

--l-

Pr0t[+21](3min)

10-

[+18B](10see)

a. oa

&

4-

o

O

20,

-2[o g, - o o. o. .o .o .o ~.

2 8II,o o, g,

T i m e (see)

-2

......

ooooo

Time (sec)

(rain)

After administration

IIJ-, . . . . . . . . o

During administration

I I , , ,

o o ~ o o o

(min) After administration

Fig. 4. Pulmonary artery systolic and diastolic pressure (PAS/PAD) changes for protamine (prot) [+21], [+ 18B] variant, and the [+ 18BE] variant, after 10-second and 3-minute administration.

L M W H reversal for all variants was noted, with [ + 18BE] at 3 minutes having equal if not greater efficiency than protamine [ + 2I] by factor Xa reversal (Fig. 5). Comparisons between groups were possible at 3 minutes and 10 minutes for factor Xa reversal with A N O V A , p < 0.01 at both time points. In considering factor Xa reversal data at 3 minutes [ + 18BE] (3 minutes) with 91% reversal was significantly superior over all four other groups: protamine [ + 2 1 ] (10 seconds) 91% versus 63%, p < 0.0i; protamine [ + 2 1 ] (3 minutes) 91% versus 68%, p < 0.05; [+ 18BE] (10 seconds)91% versus 73%, p < 0.0I; and [ + 1 8 B ] (10 seconds) 91% versus 64%, p < 0.0I. At 10 minutes after administration of [ + 18BE] (3 minutes) factor Xa reversal remained significantly improved over protamine [ + 21] (10 seconds) 68% versus 4 5 % , p < 0.01; [+ 18BE] (10 seconds) 68% versus 51%, p < 0.01; and [ + 18B] (10 seconds) 68% vs 34%, p < 0.01: These comparisons for factor Xa were not possible at 30 minutes because by A N O V A , p = 0.i56. Percent reversal for all coagulation tests was defined (Table II, Fig. 5). Platelet/leukocyte changes. Maximal thrombocytopenia for the compounds was: Protamine [ + 2i]

(10 seconds) - 4 3 % , [ + 18B] (10 seconds) - 56%, [ + 18BE] (10 seconds) - 4 4 % , protamine [ + 21] (3 minutes) - 32%, and [ + 18BE] (3 minutes) - 4 9 % at 3 minutes after reversal. Thrombocytopenia had reverted to - 8 % by 10 min and the platelet count had rebounded + 35% by 30 minutes in [ + 18BE] given over 3 minutes. Changes in platelet counts at 3, 10, and 30 minutes after compound administration are listed (Table II). In selected animals in w h o m the bleeding time was determined for both protamine [ + 2 1 ] (3 minutes) and [ + 18BE] (3 minutes), the bleeding time never surpassed 8 minutes despite the large drops in platelet counts. Maximal declines in leukocyte count were protamine [ + 21] (10 seconds) 3%, [ + 18B] (10 seconds) - 21%, [ + 18BE] (10 seconds) - 1%, protamine [ + 21] ( 3 minutes) 26%, and [ + 18BE] (3 minutes) - 7%. There were no significant differences for declines in either platelet or white blood cell counts by A N O V A between the five groups at 3, 10, or 30 minutes. Circular dichroism. The variant [ + 18BE] exhibited a spectrum typical of a random coil in water solution. The spectrum of [ + 1 8 B E ] in 100% trifluoroethanol was typical for a strong ~ helix,

J O U R N A L OF VASCULAR SURGERY Volume 21, Number 5

Wakefield et al.

100-

100-

90

90

80-

80

70-

3-- 70

60,

~0

-~ 50-

g 50

~" 4o'~

845

40

30-

}- 30

20-

20

9 Prot[+21](10sec)

10O100-

100

90 80

~

P 60

[]

[+18B1(10 sec)

[]

[+18BE](lOsec)

90

9 Prot[+21](3min)

80

[]

[+18BE](3min)

9 60 m

50

~, 40

8 40

30

30

20

3

10 Time (min)

30

3

10 Time (min)

30

Fig. 5, Percefit LMWH anticoagulation reversal as indicated by ACT, antifactor Xa activity, TCT, and antifactor IIa actiVity for pr&amin e (prot) [ § 2I], [§ 18B] variant, and [ + 18BE] variant, after 10-second and 3-minute administration.

indicating that [ + 18BE] has ~xhelix-forming capability despite the high concentration ~of positive charges (Fig. 6a). Addition of the LMWH, or even standard unfractionated heparin, to [+ 18BE] resulted in the induction of a strong c~ helix configuration (Fig. 6b). DISCUSSION

Heparin anticoagulation reversal and toxicity have been documented to correlate with the positive charges in protamine compounds. 7 The greater the net positive charge, the more toxic the compound. Commercial protamine, a mixture of highly cationic polypeptides purified from salmon sperm, has as its major species a 32-amino acid peptide with the sequence PR4SaRPVRsPRVSR6G2R48 and a total charge of [+21]. Reversal of heparin occurs as protamme binds to negatively charged heparin. This reverts antithrombin III to a conformation less effective in inhibition of factors Xa and IIa and other serine proteases in the coagulation cascade. 9,1~ Previously, we have demonstrated that acetylating and amidating a [+ 18] compound to produce [+ 18B] improved its reversal efficacy while limiting but not eliminating its toxicity. 4 We hypothesized that this

improvement was due not only to the terminal ace@ and amide groups preventing degradation by circulating aminopeptidases and carboxypeptidases, but also to an improvement in a helical-forming propensky of [ + 18B] compared with [ + 18]. The latter was attributed to an increase in the dipole moment of the c~helix formed by the [ + 18B] peptide on binding to LMWH. By circular dichroism studies, we found [ + 18BE] to be a strong a helix-forming compound on addition to LMWH. In this investigation, we enhanced the ~ helical propensity of [ + 18B] by replacing the non-~ helical-forming amino acid proline (P) by the c~ helical forming amino acid glutamic acid (E), thus forming the designer variant [ + 18BE]. By this simple change, along with the addition of one alanine for spacing considerations, we produced a compound with a TTS of 2.3 (10 seconds administration) compared with a TTS of - 3 . 1 for [+18B]. The compound [+18BE] was significantly less toxic when given over 3 minutes compared with protamine given over 3 minutes (p < 0.01) or given over 10 seconds (p = 0.02). The [ + 18BE] compound also demonstrated excellent reversal efficacy with antifactor Xa activity

JOURNAL OF VASCULAR SURGERY

846

Wakefield et al.

May 1995

/r /I !

l

/ l

.

/v

I

-30

180

_

_

..........

[+ 18BE]ia water

-

(+18BE]I~

-

I

-

I

Wavolenglh[rm]

.........

TFE

300

A

-20 180

-

LMWH(Logiparinl [+ 18BE]plusLMWH

I _

W.velen~th[nm]

300

B Fig. 6. A, Circular dichroism spectra of [ + 18BE] in water and in 100% trifluoroethanol (TFE). Strong positive deflection at approximately 190 nm and negative defection from 205 to 230 nm for [+ 18BE] in 100% TFE (solid line) are characteristic of alpha-helices. Strong negative deflection near 200 nm without other major special features (interrupted line) for [ + 18BE] in 100% water is typical for random coils. B, Circular dichroism spectra of [ + 18BE] with LMWH (Logiparin) and of LMWH alone. Relatively featureless spectrum is observed for LMWH alone (interrupted line), suggesting that it has no significant secondary structure. Circular dichroism spectrum of mixture of [ + 18BE] and LMWH exhibited all characteristic features (solid line) of alpha-helix.

being as high as 91% when administered over 3 minutes. In fact, the reversal efficiency for [ + 18BE] of 89%, 91%, 100%, and 94% at 3 minutes after compound administration for ACT, antifactor Xa, TCT, and antifactor IIa activity when administered over 3 minutes is certainly at a level where it would be adequate clinically. Importantly, reversal efficiency of [ + 1 8 B E ] remained at a level comparable to standard protamine at 10 minutes and 30 minutes after reversal for both rapid and slow rates of administration. Logiparin (LHN-1), a form of L M W H , has a molecular mass ranging from 600 to 20,000 d, with its peak being approximately 5000 d. More than 70% ranges between 1500 and 10,000 d. The biologic activity of Logiparin is approximately 87 I U / m g

antifactor Xa activity, and the antifactor Xa/antifactor IIa ratio is 1.7/1.0. Data are somewhat limited concerning Logiparin or other L M W H reversal by protamine, zl-15 However, data that do exist indicate that protamine incompletely reverses L M W H anticoagulation as assessed by antifactor Xa activity. For cardiovascular applications over and above pure prophylaxis, L M W H s have an improved pharmakokinetic profile compared with standard unfractionated heparin, decreased antiplatelet activity with potentially less bleeding risk, less lipolytic activity, and a half-life that is not doseZdependent) 6x9 Other advantages include more constant antifactor Xa inhibition, less protein C antigen decrease, less complement activation, and less inhibition ofplatelet aggregation compared with standard unfractionated

JOURNAL OF VASCULARSURGERY Volume 21, Number 5

heparin2 7 Its use has been suggested in aortic surgery and cardiopulmonary bypass procedures. 2~ We chose to study the protamine variants' effects on LMWH ant:coagulation, because there is likely to be an increase in the use of these heparins in the future. Furthermore, protamine is believed by many not to be able to fully reverse the anti-Xa activity of LMWH. We have not yet fully tested the variants [ + 18B] and [ + 18BE] and protamine against other LMWH preparations in addition to Log:par:n, although preliminary unpublished data suggest that [ + 18BE] is effective and safe when Used to reverse the anticoagulant effects of Enoxaparin. In other preliminary unpublished data from o u r laboratory, [ + I8BE] has also been found to be Very effective, with essentially no toxicity when used to reverse standard unfractionated heparin, suggesting that [ + 18BE] may be useful in reversing the ant:coagulation of multiple heparin compounds. The use of [+ 18BE] and Other re!atedpeptides for reversal of standard unfracti0nated heparin will be the subject of future investigations. A number of mechanisms have been suggested to be responsible for protamine-related toxicity including hiStamine production and release, 21-23thromboxane production, 24-29complement activation, a~ inhibition o f plasma carboxypeptidase N, 37 vasodilation of the peripheral vasculature and depression of the heart, 38 and various immunologic mechanisms. 39-47 Pulmonary artery hypertension likely is due to thromboxane release from nonplatelet sources within the pulmonary Circulation, 24,2<28,48-sa whereas systemic hypotension appears to be Caused by the elaboration of a vasoditator factor such as nitric oxide, s4-s6 along with direct depression of myocardial function. s7 Results with an isolated cardiac myocyte preparation have recently suggested that unbound protamine can enter the myocardial interstitium and inhibit myocyte contractile function and responsiveness to inotropic agents, s8 Protamine unbound, not heparin-protamine or heparin alone, was reSponsible for these adverse effects: Thus the more tightly bound the reversal agent is to heparin, the less free agent circulating is available to cause adverse responses. Thrombocytopenia and leukopenia are due to direct charge effects of protamine on phospholipid membranes.59 64 The thrombocytopenia associated with protamine or [ + 18BE], although considerable, was not associated with any prolongation of bleeding time above 8 minutes at the height of platelet count decline. For [ + 18BE] given over 10 seconds and 3 minutes, the thrombocytopenia reverted toward normal at 10 minutes, and platelet count increased above

Wakefield et al.

847

baseline at 30 minutes after administration ( + 20%, +35%) compared with protamine in which the platelet count remained depressed at 30 minutes after both rates of administration ( - 25%, -19%). This suggests another potential benefit of [ + 18BE] over standard protamine in reversal of LMWH ant:coagulation. REFERENCES 1. Wakefield TW, Lindblad B, Stanley TJ, et al. Heparin and protamine use m peripheral vascular surgery: a comparison between surgeons of the Society for Vascular Surgery and the European Society for Vascular Surgery. Eur J Vasc Surg [994;8:193-8. 2. Lindblad B, B0rgstr0m A, Wakefield TW, Whitehouse WM Jr. Stanley JC. Haemodynamic and haematologic alterations with protamine r&ersal of ant:coagulation: comparison of standard heparin and a low molecular weight heparin fragment. Eur J Vasc Surg 1987;1:181-5. 3. Montalescot G, Zapol WM, Carvalho A. Robinson DR, Torres A, Lowenstein E. Neutralization of low molecular weight heparin by polybrene prevents thromboxane release and severe pulmonary hypertension m awake sheep. Circulation 1990;82:1754-64. 4. Wakefield TW, Andrews PC, Wrobleski SK, et al. Reversal of low-molecular weight heparin ant:coagulation by synthetic protarnine analogues. J Surg Res 1994;56:586-93. 5. Ferran DS, Sobel M, Harris RB. Design and synthesis of a helix heparin-binding peptide. Biochemistry 1992;31: 5010-6. 6. Tyler-Cross R, Sobel M, Marques D, Harris RB. Heparin binding domain peptides of antithrombin III: analysis by isothermal titration calorimetry and circular dichroism spectroscopy. Protein Sci 1994;3:620-7. 7. DeLucia A III, Wakefield TW, Andrews PC, et al. Efficacy and toxicity of differently charged polycationic protamine-like peptides for heparin ant:coagulation reversal. J VASC SURG i993;I8:49-60. 8. Ando T, Watanabe S. A new method for fractionation of protamines and the amino acid sequences of salmine and three components of iridine. Int J Protein Res 1969;I: 22i-4. 9. Rosenberg RD. Chemistry of the hemostatic mechanism and its relationship to the action of heparin. Fed Proc 1977;36: 10-8. i0. Olson ST, Shore ID. Demonstration of a two-step reaction mechanism for inhibition of alpha-thrombin by antithrombin III and identification of the step affected by hepatin. J Biol Chem i982;257:14891-5. 11. Van Ryn-McKenna J, Ca: L, Ofosu FA, Hirsh l, Buchanan MR. Neutralization of enoxaparine-induced bleeding by protamine sulfate. Thromb Haemost I990;63:27i-4. 12. Lindblad B, Borgstrom A, Wakefield TW, Whitehouse WM Jr, Stanley JC. Protamine reversal of ant:coagulation achieved with a low molecular weight heparin: the effects on eicosanoids, clotting and complement factors. Thromb Res I987;48:31-40. 13. Diness V, Ostegaard PB. Neutralization of a low molecular heparin (LHN-I) and conventional heparin by protamine in rats. Thromb Haemost i986;56:318-22. 14. Racanelli A, Fareed J, Walenga JM, Coyne E. Biochemical and pharmacologic studies on the protamine interaction with

848

Wakefield et al.

heparin, its fractions and fragments. Sere Thromb ttemostas 1985;11:176-89. I5. Harenberg J, Gnasso A, de Vries JX, Zimmermann R, Augustin J. Inhibition of low molecular weight heparin by protamine chloride in vivo. Thromb Res 1985;38: 11-20. 16. HoMer E, Soderberg K, Bergqvist D, Lindahl U. Heparin and its low molecular weight derivatives: anticoagulant and antithrombotic properties. Haem0stasis i986;16(suppl 2): 1-7. i7. Mehssari E, Stringer MD, Kakkar VV. The effect of a bolus injection of unfractionated or low molecular weight heparin during aortobifemoral bypass grafting. Eur J Vase Surg 1989;3:I21-6. 18. Kroneman H, Eikelboom BC, Knot EAR, et al. Pharmaeokinetics of low-molecular-weighther~arinand unfractionated heparin during elective aortobifemoral bypass grafting. }"VASCSURG 1991;14:208-14. 19. Wilson NV, Melissari E, Standifield NJ, Kakkar VV. Intraoperative antithrombotic therapy with low molecular weight heparin in aortic surgery: How should heparin be administered? Eur J Vase Surg 199I;5:565-9. 20. Koza MJ, Messmore HL, WallockME, Walenga JM, Pifarre R. Evaluation of a low molecular weight heparin as an anticoagulant in a model of cardiopulmonary bypass surgery. Thromb Res 1993;70:67-76. 21. Horrow JC. Protamine: a review of its toxicity. Anesth Analg 1985;64:348-61. 22. Keller R. Interrelations between different types of cells: II. Histamine-release from the mast cells of various species by cationic polypeptides of polymorphonuclear leukocyte lysosomes and other cationic compounds. Int Arch Allergy Immunol 1968;34:139-44. 23. Tobin MC, Karns BK, Anselmino LM, Thomas LL. Potentiation of human basophil histamine release by protamine: a new role for a polycation recognition site. Mol Immunol 1986;23:245-53. 24. Morel DR, Zapol WM, Thomas SJ, Kitain EM, Robinson DR, Moss J, Chenoweth DE, Lowenstein E. CSa and thromboxane generation associatedwith pulmonary vaso- and broncho-constriction during protamine reversal of heparin. Anesthesiology 1987;66:597-604. 25. Conzen PF, Habazettl H, Gutmann R, et al. Thromboxane mediation of pulmonary hemodynamic responses after neutralization of heparin by protamine in pigs. Anesth Analg 1989;68:25-31. 26. Degges RD, Foster ME, Dang AQ, Read RC. Pulmonary hypertensive effect of heparin and protamine interaction: evidencefor thromboxane B2 releasefrom the lung. Am J Surg 1987;154:696-9. 27. Hobbhahn J, Conzen PF, Zenker B, Goetz AE, Peter K, Brendel W. Beneficialeffect of cyclooxygenaseinhibition on adverse hemodynamic responses after protamine. Anesth Analg 1988;67:253-60. 28. McIntyre RW, Flezzani P, Knopes KD, Reves JG, Watkins WD. Pulmonary hypertension and prostaglandins after protamine. Am J Cardiol 1986;58:857-8. 29. Morel D, Lowenstein E, Nguyenduy T, et al. Acute pulmonary vasoconstriction and thromboxane release during protamine reversal of heparin anticoagulation in awake sheep: evidencefor the role of reactive oxygenmetabohtes following nonimmunologic complement activation. Circ Res 1988;62: 905-15.

JOURNAL OF VASCULARSURGERY May 1995

30. Cavarocchi NC, Schaff I-IV, Orsulak TA, Homburger HA, SchnetlWA Jr, Pluth JR. Evidencefor complement activation by protamine-heparin interaction after cardiopulmonary bypass. Surgery 1985;98:525-3I. 31. Fehr J, Rohr H. In vivo complement activation ofpolyanionpolycation complexes: evidence that C5a is generated intravascularlyduring heparin protamine interaction. Clin Immunol Immnnopathol 1983;29:7-14. 32. Kirklin JR. In discussion: Shapira N, SchaffHV, Piehler IM, White RD, Sill JC, Pluth JR. Cardiovascular effects of protamine sulfate in man. J Thorac Cardiovase Surg 1982; 84:513. 33. Kirklin j'K, Chenoweth DE, Naftel DC, et al. Effects of protamine administration after cardiopulmonary bypass on complement, blood elements and the hemodynamic state. Ann Thorae Surg 1986;41:193-9. 34. Rent R, Ertel N, Eisenstein R, Gewurz H. Complement activation by interaction of polyanions and polycations: I. Heparin-protamine induced consumption of complement. J immunol 1975;114:120-4. 35. SiegelI, Rent R, Gewurz H. Interactions of C-reactiveprotein with the complement system: I. Protamine-induced consumption of complement in acute phase sera. J Exp Med i974; 140:631-47. 36. White JV. Complement activation during cardiopulmonary bypass. N Eng J Med 1981;305:5I [Letter]. 37. Tan F, Jackman H, Skidgel RA, Zsigmond EK, Erdos EG. Protamine inhibits plasma carboxypeptidaseN, the inactivator of anaphylatoxins and kinins. Anesthesiology 1989;70: 267-75. 38. Goldman BS, Joison I, Austen WG. Cardiovasculareffects of protamine sulfate. Ann Thorac Surg I969;7:459-71. 39. Best N, Teisner B, Brudzinskas JG, Fisher MM. Classical pathway activation during an adverse response to protamine sulphate. Br J Anaesth 1983;55:1149-53. 40. Caplan SN, Berkman EM. Protamine sulfate and fish allergy. N Engl J Med 1976;295:I72 [Letter]. 41. Doolan L, McKenzie I, Krafchek J, Parsons B, Buxton B. Protamine sulphate hypersensitivity. Anaesth Intensive Care 1981;9:I47-9. 42. Knape JA, SchullerJL, de-Haan P, de-Jong AP, BoviUJG. An anaphylacticreaction to protamine in a patient allergicto fish. Anesthesiology 1981;55:324-5. 43. Lakin JD, BlockerTJ, Strong DM, Yoeum MW. Anaphylaxis to protamine sulfate mediated by a completion-depleted IgG antibody. J Allergy Clin Immunol i978;651:102-7. 44. Levy JH. Life-threatening reactions to intravenous protamine. N Eng J Med 1989;321:1684 [Letter]. 45. LevyJH, Zaiden JR, Faraj B. Prospectiveevaluation of risk of protamine reactions in patients with NPH insulin-dependent diabetes. Anesth Analg 1986;65:739-42. 46. LevyJH, SehwiegerIM, Zaiden JR, Faraj BA, Weintraub WS. Evaluation of patients at risk for protamine reactions. J Thorac Cardiovasc Surg 1989;98:200-4. 47. Weiss ME, Nyhan D, Peng ZK, et al. Association of protamine IgE and IgG antibodies with life-threatening reactions to intravenous protamine. N Engl J Med i989;320: 886-92. 48. Cheng SW, Westcott JY, Hensen JE, VoelkelNF. Pulmonary vascular injury by polycations in perfused rat lungs. J Appl Physiol I987;62:1932-43. 49. Jastzebski J, Hilgard P, Sykes MK. Pulmonary vasoconstriction produced by protamine and protamine-heparin complex

JOURNAL OF VASCULARSURGERY Volume 21, Number 5

50. 51. 52.

53.

54. 55.

56.

57.

in the isolated cat lung perfused with blood or dextran. Cardiovasc Res 1975;9:691-6. Lowenstein E, Johnston WE, Lappas DG, et al. Catastrophic pulmonary vaso-constriction associated with protamine reversal of heparin. Anesthesiology 1983;59:470-3. Montalescot G, Kreil E, Lynch K, et al. Effect of platelet depletion on lung vasoconstriction in heparin-protamine reactions. J AppI Physiol 1989;66:2344-50. Olinger GN, Becker RM, Bonchek L1. Noncardiogenic pulmonary edema and peripheral vascular collapse following cardiopulmonary bypass: rare protamine reaction? Ann Thorac Surg 1980;29:20-5. Wakefield TW, Bouffard JA, Spaulding SA, et al. Sequestration of platelets in the pulmonary circulation as a consequence of protamine reversal of the anticoagulant effects of heparin. J VASCSURG 1987;5:187-92. Akata T, Yoshitake J, Nakashima M, Itoh T. Effects of protamine on vascular smooth muscle of rabbit mesenteric artery. Anesthesiology 1991;75:833-46. Ignarro LJ, Gold ME, Buga GM, et al. Basic polyamino acids rich in arginine, lysine, or ornithine cause both enhancement of and refractoriness to formation of endothelium-derived nitric oxide in pulmonary artery and vein. Circ Res 1989;64: 315-29. Pearson PJ, Evora PR, Ayrancioglu K, SchaffHV. Protamine releases endothelium-derived relaxing factor from systemic arteries: a possible mechanism of hypotension during heparin neutralization. Circulation 1992;86:289-94. Wakefield TW, Bies LE, Wrobleski SK, Boiling SF, Stanley JC, Kirsh MM. Impaired myocardial function and oxygen utilization due to protamine sulfate in an iso-

Wakefield et al.

58.

59. 60. 61. 62.

63.

64.

849

lated rabbit heart preparation. Ann Surg 1990;212: 387-94. Hird RB, Crawford FA Jr, Mukherjee R, Zile MR, Spinale FG. Effects of protamine on myocyte contractile function and beta-adrenergic responsiveness. Ann Thorac Surg 1994;57: 1066-75. Eika C. On the mechanism ofplatelet aggregation induced by heparin, protamine and polybrene. Scand I Haematol 1972; 9:248-57. lacques LB. A study of the toxicity of the protamine, salmine. Br I Pharmacol 1949;4:135-44. Lindblad B, Wakefield TW, Whitehouse WM lr, Stanley I t . The effect of protamine sulfate on platelet function. Scand J Thorac Cardiovase Surg 1988;22:55-9. Wakefield TW, Hantler CB, Lindblad B, Whitehouse WM Jr, Stanley JC. Protamine pretreatment attenuation of hemodynamie and hematologic effects of heparin-protamine interaction: a prospective randomized study in human beings undergoing aortic reconstructive surgery. J VAse SURG 1986;3:885-9. Wakefield TW, Whitehouse WM Jr, Stanley JC. Depressed cardiovascular function and altered platelet kinetics following protamine sulfate reversal of heparin activity. J VAse SURG 1984;1:346-55. Wakefield TW, Lindblad B, Whitehouse WM lr, Hantler CB, Stanley }'C. Attenuation of hemodynamic and hematologic effects of heparin-protamine sulfate interaction after aortic reconstruction in a canine model. Surgery 1986;100:45-50.

Submitted Sept. 27, 1994; accepted Nov. 28, 1994.

DISCUSSION

Dr. D o n a l d Spadone (Columbia, Mo 0. This study is part of the continuing evolution of a bioengineered protamine variant developed by Dr. Wakefield and his group in an effort to decrease the toxicity and increase the efficacy of reversal of unfractionated heparin and L M W H . They're on their way to these small, low-molecular-weight fractions. Only one L M W H is available in the United States for the use of prophylaxis against deep venous thrombosis and pulmonary embolization in patients undergoing hip and knee prosthetic replacement. There are multiple other low-molecular-weight preparations that will soon be available. Because of a mixture of different polysaccharide lengths, each of these preparations has a slightly different half-life, antifactor Xa, and antithrombin efficiency. L M W H fractions containing less than 18 polysaccharide units continue to bind antithrombin-III and complex effectively to inhibit antifactor Xa. These L M W H and antithrombin-III complexes cannot, however, inhibit thrombin. The potential advantage of the use of L M W H is that they have a longer plasma half-life. They can be administered once a day. They have a better bioavailability than standard unfractionated heparin. The clinical potential for these low-molecular weight agents is for better pro-

phylaxis for panents at high risk who have had an acute stroke or have had spinal cord injury. In clinical studies in patients who have an established deep venous thrombosis, L M W H has been as safe and efficacious as traditional intravenous unfractionated heparin preparanons. This agent may also prove to be effective in short-term anticoagulation needed for hemodialysis, cardiopulmonary bypass, and intraoperative arterial reconstrucnve procedures. In this report Dr. Wakefield and his group have shown that their protamine variant species is more effective than commerciallv available protamine to reverse L M W H and has significantly less toxmity in this animal model. I believe this is a significant advance because of the decreased toxicity and the ability to better reverse the antifactor Xa effect. Because different L M W H preparations have different polysaccharide combinations and polysaccharide lengths, have you checked your variant protamine against other low-molecular-weight preparations? Is your 18BE as efficacious for neutralization of these other preparations? What is the long-term toxicity of this designer protamine variant? Are there any temporary or permanent conformational changes that occur to the anti-thrombin III L M W H

850

Wakefield et al.

complex on interaction with this new designer protamine variant? The authors have studied reversal of anticoagulation when heparin was administered intravenously. In the future, L M W H undoubtedly will be given through the subcutaneous route. Have you studied reversal of L M W H that was given via the subcutaneous administration. They also studied reversal only up to 30 minutes after the administration of their protamine with intravenous administration of LMWH. Is there a late rebound effect of anticoagulation after the initial reversal with 18BE? Why is there a difference in toxicity that appears to be related to the rate of protamine infusion? Will this designer variant of protamine cause the same allergic response in patients with diabetes who have been previously sensitized to protamine found in preparations such as NPH-insulin? H o w soon can we expect to see trials of this agent on human beings? Dr. Thomas W. Wakefield. Regarding the first question you asked, we have some preliminary results with enoxaparin and the [ + 18BE] compound and it appears also to be effective against this particular L M W H with little toxicity. However, we have not performed enough experiments to be able to make firm conclusions about this or other LMWHs. We do not know about any long-term toxicity of these designer variant peptides. However, we do not expect that there would be any major long-term toxicity. We do not believe there are any temporary or permanent conformational changes that would occur to the antithrombin-III L M W H complex upon interaction with [ + 18BE] because the TCT, a measure of thrombin activity, reverses 100% with the administration of this agent for both L M W H and standard unfractionated heparin. We have not studied subcutaneous administration of

JOURNALOF VASCULARSUKGERY May 1995

heparin and reversal by our variants, although I would not expect it to be any different. The heparin circulating will be reversed no matter whether it is given intravenously or subcutaneously. We have only studied subjects for 30 minutes after peptide administration. We have not found any heparin rebound at 30 minutes. It would be interesting to investigate for rebound later on, but we do not expect any late heparin rebound. Why does toxicity appear related to the rate of protamine infusion? We have always believed that this is related to the fact that whatever factors released are responsible for the adverse responses (nitric oxide for the hypotension and thromboxane for the pulmonary hypertension, for example) will be released much more rapidly if a large bolus of protamine is given rather than slower administration. In our model, we have tested very rapid rates of administration to magnify the adverse responses, and it can clearly be stated that if toxicity is not seen with 10 seconds of administration, it should not be seen with 3 minutes or longer administration. The designer variants should not cause allergic responses in patients with diabetes if they have n o t been exposed to them before. They are a different structure than protamine, although on the same general backbone structure as protamine, with different amino acids. Unless a patient has been exposed to them in the past, there should be no allergic potential. H o w soon will these variants be available? Further developments and refinements will be necessary before they will become available clinically, but we hope they will be available in the near future.