J Mol Cell Cardiol22, 1325-1335 (1990)

Lactate

Does Not Enhance A.noxia/Reoxygenation Rat Cardiac Myocytes Timothy

Department

P. Geisbubler

of Physiology, University of Missouri,

and Michael MA USA

Damage

In Adult

J. Rovetto*

415 Medical

Sciences, Columbia,

MO 65212,

(Received 30 June 1989, accepted in revised form 26 June 1990) T. P. GEISBUHLER AND M. J. ROVETTO. Lactate Does Not Enhance Anoxia/Reoxygenation Damage In Adult Rat Cardiac Myocytes. Journal ofMoleculurandCellular Curdiology (1990) 22, 13251335. Accumulation oflactate in myocardial cells has been proposed as a primary trigger ofischemic damage in heart. This hypothesis was tested using isolated cardiac myocytes from adult rats. Cells were subjected to anoxialreoxygenation protocols in the presence or absence of lactate at two extracellular pH values. Reductions in total rods and increased numbers of shortened rods (“contracted” cells) were evident in cell populations exposed to anoxia and in reoxygenated populations in the absence of glucose at pH values of 6.9 and 7.3. Although lower pH reduced cell adenine nucleotide contents over those seen at higher pH, neither 10 rnhr nor 50 mM lactate enhanced nucleotide loss or caused extra morphological damage under any condition in this study. Therefore, under conditions simulating those in ischemic heart, cell damage could not be attributed to high lactate concentrations.

Introduction Myocardial ischemia is characterized by a reduction or interruption of blood flow to the cardiac muscle, such that sufficient oxygen no longer reaches the tissue and metabolic endproducts are not washed out. The reduced contractility and performance of the affected tissue are well documented; the biochemical causesof these phenomena are lessclear. Coronary occlusion causes an immediate shift from aerobic to anaerobic metabolism in cardiac tissue (reviewed in Jennings and Reimer, 1981), resulting in (among other things) accumulation of lactate and protons as metabolic endproducts (Williamson et al., 1976). Accumulation of these products has been advanced by some researchersas a primary trigger for irreversible ischemic damage (Neely and Grotyohann, 1984). Our study was designed to test this idea, using acutely isolated myocytes as a test system. Isolated myocytes can be prepared so that most of the important morphologic and biochemical characteristics of the adult cell in vivo are retained. Moreover, the morphologic progression of these cells from “health” to *To

whom

0022-2828/90/l

reprint

requests

should

11325 + 11 $03.00/O

be sent at the above

“death” during anoxia or reoxygenation following anoxia is well established (Brierley et al., 1986). Isolated myocytes therefore represent a powerful tool for determining the sequence of biochemical events preceding observable anoxic damage. In this study, we exposed myocytes to brief periods of anoxia so that the cells were reversibly damaged. Because cardiomyocytes are permeable to both lactate and proton via a specific translocase(Dennis et al., 1984; Kammermeier et al., 1985; Trosper and Philipson, 1987), addition of either or both of these metabolic endproducts to the external medium should generate excessmortality in the cell population, if either endproduct triggers anoxic damage. Our data suggest that lactate alone is insufficient to generate damage in these cells. Methods Isolation of myocytes

Myocytes were isolated from adult rat hearts as described previously (Geisbuhler et al., 1987). Four male Sprague-Dawley rats (250address. 0

1990 Academic

Press Limited

1326

T. P. Geisbuhler

and M. J. Rovetto

350 gj were injected with heparin (200 IU, ip) 20 min before anesthesiawith sodium pentobarbital (105-140 mg/kg, ip). The hearts were excised and placed in ice-cold minimal essential medium (Joklik-modified) buffer containing the following additional compounds (in mM): 60 taurine, 20 creatine, and 5 HEPES. (This mixture is hereafter called buffer, and is nominally calcium-free unless otherwise stated.) The aortas were cannulated with 2mm ID stainlesssteel tubing and perfused at 40 mmHg hydrostatic pressure.The perfusate [buffer supplemented with 0.1 o/0 bovine serum albumin (BSA) and 1 IU heparinlml] was equilibrated with 100% 02 at 37°C (pH = 7.3) and did not recirculate. After a 5 to 10 min washout period, the perfusate was changed to buffer containing 0.7 mg collagenase/ml (classI). This was allowed to recirculate until the aortic perfusion pressure decreasedbelow 40 mmHg (30-40 min) . The ventricles were cut from the hearts, minced, and placed in fresh collagenase-containing perfusate which was maintained in equilibrium with 100% OZ. The tissuewas shaken 10 min at 250 rpm in a New Brunswick model G-76 gyrotory water bath (New Brunswick Scientific, Edison, NJ). CaCls (50 PM) was added to the minced tissue and digestion continued for 10 min. The cells were dispersed, filtered through a double layer of cheesecloth, and diluted 1:4 with buffer containing 0.1% BSA. The cells were settled, the supernatant removed, and the cells resuspended in buffer containing 1.5% BSA. (The majority of the cells in the supernatant were hypercontracted or permeable to trypan blue, and were removed). The resuspension-settling procedure was repeated twice, and the cells were suspended in buffer containing 1.5% BSA. At this point in the isolation procedure, the cell preparation was essentiallyfree of nonmuscle cardiocytes. The cells were continuously aerated with moist lOOo/0 02 and maintained in suspensionat 37°C by gentle swirling. Calcium chloride was added in lOO200 pM stepsover 30 min to a final concentration of 1 mM. Finally, the cells were settled to remove additional hypercontracted cells from the preparation, and the cells in the pellet suspendedin buffer containing 1.5% BSA and 1 mM CaC12. The cell suspensionwas exam-

ined for rod-shaped cells and viability using 10% glutaraldehyde/0.3% trypan blue as described by Hohl et al. (1983). The myocyte preparations routinely contained 70-90% rod shaped cells and >90% viable cells. Studies were completed within 3 h of the final addition of CaClz. General experimental protocol

All incubations were conducted at 37°C in siliconized glass flasks, stoppered and gassed continuously with O2 or Nz as appropriate. Buffers used in the incubations were the same as described above under Isolation of Myocytes and contained 1.5% BSA and 1 mM CaCla. These buffers were equilibrated at least 1 h with either 100% 02 or 100% N2. Four 200 ~1 aliquots were removed at the time intervals indicated in a given protocol. One was fixed and stained for examination of cell viability and morphology, and the other three were extracted and neutralized for nucleotide analysis. Myocytes from experimental protocols were fixed and stained using 10% glutaraldehyde/ 0.3% trypan blue as described by Hohl et al. (1983) and examined by light microscopy. Becausemorphology was used as an indicator of reoxygenation damage, careful note was taken of both membrane integrity and gross morphology. Cells were classed as viable if they excluded trypan blue dye (unstained); those unable to exclude the dye (stained blue) were non-viable. Myocytes were further classedas either rod-shaped (exhibiting normal morphology, length/width = 8-lo), contracted (more box-shaped, length/width= less than 3), or hypercontracted (amorphous rounded shape). Examples of these different morphologies are shown in Figures 1 a-c. The data from cell counts are presented in Figures 2 through 5. The data representedin figures as“rods” do not include the more shortened “contracted” forms. Cells which were contracted to the shortened morphology described above are designated “contracted”. Viability, discussedabove, is designated “viable”. Cell aliquots used for nucleotide analysis were quickly transferred to 1.5 ml microcentrifuge tubes containing (top to bottom)

FIGURE I. Demonstration of anoxic and reoxygenation damage in cardiac myocytes. Cells were incubated at 37°C in siliconized glass flasks, either oxygenated [Panel (a)] or anoxic [Panel (b)] as described in Methods. After removing aliquots for cell counting and nucleotide analysis, the anoxic cells were exposed to oxygen for 4 min, and an aliquot removed for cell counting [Panel (c)l. Typical rod-shaped myocytes (rod), partially damaged cells (con) and hypercontracted cells (md) are labeled, and correspond to those described in Methods. Photographs were taken using a Nikon Labophot light microscope and Kodachrome 64 slide film. Magnification x 100 (bar = 100 pmi.

T. P. Geisbuhler and M. J. Rovetto

1328 ( 0 ) Reoxygenoiion -e-

20 IO

domoge

Rods

in myocytes

-o-

oxygenated

Viable

-I-

controls

Contracted

/ _._._.

-..

0

-.-.-

.-..

I5

_-.-.

45

A~OX~O for time

(bl -a100 90

Rods

__~_._._._

30

-o-

---------p-----

VlOble

---.

60

lndlcated -.-

9-------

Contracted

0.7 ml ice-cold inhibitor-stop solution satuS-6-(bnitrobenzyl)thioinosine rated with (NBMPR, a nucleoside transport inhibitor), 0.4 ml bromododecane (oil), and 0.1 ml 2 M perchloric acid. The tubes were immediately centrifuged at 13000 x g for 30 s. The aqueouslayer above the oil was removed, the tube above the oil was washed, and the extract was diluted and neutralized as previously described (Geisbuhler et al., 1987). The neutralized extract was frozen at -70°C until assayedfor nucleotides. The precipitated cell pellet was disolved in 0.5 MNaOH by heating in a boiling water bath. Protein content was estimated as described by Lowry et al. (1951) using BSA as a standard.

----------.,,

Time course of cell damage 60

-

50

-

40

-

30

-

20 IO

c___Y--~ -..

0

I5

(c)

Reox --+-

30

for 4 minutes

Rods

-e-

45

at t!me

Vmble

60

Indlcoted -.-

contracted

50 40 30 20

T

t

\

IO 0

I5 Time

30 (minute)

45

60

FIGURE 2. Demonstration of anoxic and reoxygenation damage in cardiac myocytes. Cells were incubated at 37°C in siliconized glass flasks, either oxygenated [Panel (a)] or anoxic [Panel (b)] for 0, 15, 30, or 60 min as described in Methods. After removing aliquots for cell counting and nucleotide analysis, the anoxic cells were exposed to oxygen for 4 min, and an aliquot removed for cell counting [Panel (c)]. Rods, contracted cells, and viable cells correspond to those described in Methods. Values are the mean f S.E.M. of 3 experiments (each individual mean being the average of at least 3 determinations), expressed as a percentage of total cells counted.

At T = 0 min, the cells were settled and the buffer replaced with either aerobic or anoxic buffer. The cells were placed into flasks, stoppered securely, and a steady stream of either 02 or Nz applied for the duration of the incubation. After 15, 30, and 60 min incubation, aliquots were removed for cell counting and nucleotide analysis. After the last sample was taken, oxygen was reintroduced to the anoxic cell suspensions.After 4 min exposure to oxygen, another aliquot was removed for cell counting.

Effect of pH andglucose One “experiment” consisted of four separate flasks: (1) oxygenated (100% 0,) for 60 min, so that any container effect could be observed; (2) anoxia ( 100% N2) for 30 min, followed by reoxygenation (loons 0,) for 30 min; (3) 30 min anoxia/30 min reoxygenation with 10 mM added lactate; (4) 30 min anoxial30 min reoxygenation with 50 mM added lactate. This experimental protocol was designed to reveal any effect that lactate might have on anoxic or reoxygenation damage. Glucose and pH were then modified within this basic protocol. These modifications were: (1) 11 mM glucose, pH = 7.3; (2) no glucose, pH = 7.3; (3) no glucose, pH = 6.9. The final cell suspensionwas divided into four equal parts of 5 ml (cu. 5 mg cell protein/

Damage (0)

=” s i3 P % $?

in Rat Cardiac

Anox~o XI m,nu+es, T = 30 mmutes ph,,,= 7.3. Glucose = II mM

cn 5 ” G 2 5 ,-” N7

Cc)

Rods Adenlne

o

NzNOLAC

0 and

h

N*

“ET

Vfoble

nucleolldes

90 00 70 . 60 50 40. 30 20. IO-

NZ

‘YTC” -

1329

( b 1 Reoxygenotlon 30 mmutes, T = 60 rn~nute~ OH,,, = 7.3, Glucose = I I rn~

1001 90 00 70 60 50 40 30 20 IO NeNOLAC

myocytes

N:2 50 “h LAS

‘72

Contracted

td

IMP

i

Rods

0

Vloble

Adentne

nucleotldes

0 ond

:

1

Contracted IMP

I

LAC Lyrl (e I c 2.00 E6 1.60 I.00 = z f , :, $

ATP

W Guomne

Total

AN

I

-

IMP

ATP

0 Guonlne

0

GTP

If)

nucleotldes

Totol AN nucleotfdes

LAC Cl

IMP

d z

1.40 1.20 1.00 0.00 0.60 0.40 0.20

FIGURE 3. Cardiac myocyte viability and morphology in the presence of lactate, I1 mM glucose, pH = 7.3. Cells were anoxic from O-30 min [Panels (a), (c) and (e)], then reoxygenated after samples were taken for cell counting and nucleotide analysis [Panels (b), (d) and (f)]. Values are the mean f S.E.M. of four experiments, expressed as a percentage of total cells counted [Panels (a) and (b)] or as nmol nucleotide/mg cell protein [Panels (c) through (f)]. For comparative purposes, starting values (T = 0 min) for Figure 3 were: rods = 85 + 4O/,, viable = 92 + 4%, contracted = 1 f 1%; ATP = 22.2 f 0.7, total adenine nucleotides (TAN) = 27.0 + 0.6, IMP = 0.7 + 0.2 nmol/mg cell protein; GTP = 0.92 + 0.06, total guanine nucleotides (TGN) = 1.67 _+ 0.08 nmol/mg cell protein. l = significantly different (P < 0.05) from aerobic controls within the same time group.

ml) and allowed to settle for about 3 min in plastic 10 ml graduated cylinders. Each cell pellet was washed twice with the buffer in which the cells would be incubated. Lactate was added to a final concentration of 10 mMor 50 mM in two of the incubation flasks. Differences in osmolarity were eliminated by the addition of NaCl, such that lactate + NaCl in

each flask equaled 50 mM. The experiments were initiated by placing aliquots of cells into flasks; the flasks then were stoppered securely and incubated at 37°C. Flasks were supplied continuously with 0s or N2 for the duration of the experiment. At 0, 30 (pre-reoxygenation), and 60 min, 4 x 200 ~1 aliquots of cell suspension were removed; one was fixed in lo‘+/;,

T. P. Geishuhler

330 (a

)

Anoxla

30mmutes, pHou, = 7.3,

and M. J. Rovetto

T = 30 *mutes no glucose

( b ) Reoxygenatlon pH,,,=

1001

E $ z 5 T5 ,\”

30 minutes, T = 60 minutes 7.3, “a glucose

1

90 80 70 60 50 40 30 20 IO n

90

80 70

=” 2 G g “0

Aerobic

N2

N?NOLAC

N2

J

60

50 40 30 IO

200 1-

8

Aerobtc-

YzNOLAC

N2 IO I& LAC

‘T -

Rods

-

Adenlne

nucleotldes

cc I

-

301 : z h = e

Viable

0 and

Contracted IMP

-

/ E E

25

I5 IO

; z

5 0

Aerobtc

iB’d

ATP

(e) 2.00 I I.80 I.60 1.40 -

NpNOLAC

N? 10 mht LAC

-Total Guamne

AN

Rods

0

Ademne

nucleothdes

Contracted IMP

I5 IO

5 0

IMP

Aerobw

(f

nucleotldes

u and

20

N,NOLAC

50;nM LLIC -

Viable

25

: = G 0 T 6 P

20

P \

(d) --

N2 50iM LAC

ATP

1

NP 10 R;M LAC

@Z.z Total AN Guomne nucleotldes

LAC =

IMP

c

1.20 1.00 0.80 0.60 0.40 -

a20 0.00

-

N2 IO mM LAC 0

N?

50mM LAC Total GN

Aerobic

N2 NO LAC

N2 '",Z,"

0

GTP

u

N2 WmM LAC Total GN

FIGURE 4. Cardiac myocyte viability and morphology in the presence of lactate, no glucose, pH = 7.3. Cells were anoxic from (t30 min [Panels (a), (c) and (e)], then reoxygenated after samples were taken for cell counting and nucleotide analysis [Panels (b), (d) and (f)]. Values are the mean + S.E.M. of four experiments, expressed as a cell protein [Panels (c) through (f)]. For percentage of total cells counted [Panels (a) and (b)] or as nmol nucleotide/mg comparative purposes, starting values (T = 0 min) for Figure 4 were: rods = 88 f l”/0, viable = 94 + 4%, contracted = 0 f 0%; ATP = 21.9 f 1.0, total adenine nucleotides (TAN) = 28.0 f 1.3, IMP = 2.8 + I .5 nmol/mg cell protein; GTP = 1.09 + 0.09, total guanine nucleotides (TGN) = 1.76 ?r: 0.04 nmol/mg cell protein. * = significantly different (P < 0.05) from aerobic controls within the same time group; * = significantly different (P < 0.05) from anoxic controls within the same time group.

glutaraldehyde/0.3% ation of morphology nucleotide analysis.

trypan blue for examinand three were used for

Nucleotide analysis Neutralized cell extracts were analyzed for 5’nucleotides as described by Geisbuhler et al. (1984), using a Whatman SAX lo/25 PXS column.

instruments and reagents Lactate used in this study was purchased as the free acid from Sigma Chemical Co. (St. Louis, MO), and dissolved in sufficient NaOH to achieve a pH of 6.9 or 7.3. Stock solutions of lactate were prepared within 2 h of the beginning of each experiment to minimize the amount of polymerized lactate. Minimal essential medium (Joklik-modified) was pur-

Damage

in Rat Cardiac

90 =“3 80 z 70 a 60 ‘0 50 b 40

=” d E 5 z

30 20

08

IO G nerotsc

myocytes Reorygenollon 30 mlnufes, T = 60 pH,,“,= 6.9, “0 QlUCO% (b) 100 -90 70 60 50 40 30

20 IO 0

Ns

Aerobic L

lOk.4

LAC (c

\

IO

2 22 z

5

0

Adenlne

Viable

n

n~cleo+~des

and

LAC Cantrocted

LAC (d

IMP

AerObrC

-

N2NOLAC

ATP

1

Guanme

N2 10 rnM LAC loto!

251

k 5” s” \

20

N2 50mM LAC

AN

-

1 301

i

zP

0

(e

Rods

I

; 0 Fi r;;

Rods

0

0

Vlob’e

Aden,ne

r~ucleorldes

ord

If I 2.00 1.80

ATP

I I WC

0

N2 IO mM LAC

-

Total

N2 50 rnM LAC

Tot01 AN

cz-

N2

N2 50 rnM iac

IMP

-

i

Perooit

N2

NO

LAC

IO nlM LAC GTP

--1

I I

IO 5 0 ACrOblC NzNU

1.60 1.40 I.20 I.00 0.80 0.6 0 0.40 0.20

‘td

IMF

15

0.00

0

N, 50‘mM LAC Contrn,-

-

IMP

nucieor~des

minutes

80

g N,NOIAC

1331

GN

FIGURE 5. Cardiac myocyte viability and morphology in the presence of lactate, no glucose, pH = 6.9. Cells were anoxic from O-30 min [Panels (a), ( c ) and (e)], then reoxygenated after samples were taken for cell counting and nucleotide analysis [Panels (b), (d) and (f)]. Values are the mean f S.E.M. of four experiments, expressed as a percentage of total cells counted [Panels (a) and (b)] or as nmol nucleotide/mg cell protein [Panels (c) through (f)], For comparative purposes, starting values (T = 0 min) for Figure 5 were: rods = 87 + 294, viable = 95 _+ 2. contracted = 2 + 196; ATP = 15.3 + 2.0, total adenine nucleotides (TAN) = 19.1 + 2.4, IMP = I. 1 k 0.5 nmol/mg cell protein; GTP = 0.49 + 0.05, total guanine nucleotides (TGN) = 0.92 f 0.21 nmol/mg cell protein. * = significantly different (P < 0.05) from aerobic controls within the same time group.

chased from Grand Island Biological Co. (Grand Island, NY). Bovine serum albumin (BSA; Cohn fraction V, not fatty acid free) was obtained from Miles Diagnostics Division (Kankakee, IL) and dialyzed 24 h against salt solution before use (Geisbuhler et al., 1987). Collagenase was obtained from Worthington Biochemical Co. (Freehold, NJ) and tested by lot number for suitability before use. NBMPR was purchased from Aldrich Chemical Com-

pany (Milwaukee, WI). All other chemicals used were of the highest grade commercially available. Statistics

Significance between meansshown in Figures 2(a) through 5 (f) was established using one way analysis of variance, and cross-checked with Student’s t-test.

1332

T, P. Geiduhler

Results

These experiments were designed to determine if lactate enhanced anoxic or anoxia/ reoxygenation damage of myocardial myocytes. This hypothesis was tested by inducing mild damage in isolated cardiac myocytes by exposure to anoxia, and adding neutralized lactate as an experimental variable. If lactate alone is a component of the cellular response to anoxia or the oxygen paradox, cell damage and mortality should be increased in incubations with high lactate concentrations. We first determined that reoxygenation damage could be induced in isolated myocytes, and estimated the amount of exposure to anoxia that produced “mild” damage. The results of these experiments are shown in Figures 1 and 2. Figure 1(a) showscells after 60 min aerobic incubation without glucose. (Glucose must be omitted from the medium in order to observe anoxic or reoxygenation damage). These cells are obviously undamaged, showing no effects of either buffer or incubation. Figure l(b) shows cells which have been exposed to 60 min anoxia. These cells are clearly different from the cells in Panel (a), showing a dominance of the shortened forms (“contracted”, designated CON in the photo) which we find are associated with extended periods of anoxia. Figure 1(c) shows cells reoxygenated for 4 min after 60 min anoxia. A large fraction of these cells is obviously damaged, as evidenced by the rounded appearance of the cells. Two min reoxygenation was insufficient to generate this effect (data not shown). Longer periods of reoxygenation did not increasethe percentage of damaged cells. Figure 2 represents in numerical form the photographic evidence of Figure 1. Again, Panel (a) represents cells incubated aerobically for 60 min. Contracted cells did not increase during the incubation (2% at 0 min vs. 5% at 60 min), nor did rods or viability significantly decrease(95% at 0 min vs. 85% at 60 min). No significant effect of either buffer or container was observed in this system. Exposure of the myocytes to 60 min anoxia [Fig. 2(b)] damaged a significant number of the cells (contracted or hypercontracted cells). Only 39% of the cells were undamaged rods

and M. J. Rwetco

after this treatment (as opposed to 75% rods in the oxygenated control and 85% rods at 0 min). An increasein the number of contracted cells (43% at 60 min vs. 2% at 0 min) accounted for virtually all of the decline in rod-shaped cells. Reoxygenation for 4 min damaged almost all undamaged cells in the preparation [Fig. 2(c)]; only 7% of the cells were rod-shaped after this treatment. Most of the damaged cells (80%) in Figure 2(c) were hypercontracted but viable. A smaller number (13%) were contracted cells. In contrast to the situation at 60 min, 30 min anoxia damaged only a small fraction of the myocyte population [Fig. 2(b)]. Most of the cells in the population were still rods after 30 min of anoxia (72% vs. 77% in the oxygenated control and 85% at 0 min). Although contracted cells were increased by this treatment (12%), they did not represent a major percentage of the cells at this time point. Reoxygenation of these cellsfor 4 min slightly decreasedrod-shaped (66%) and contracted cells (9%) compared to the anoxic preparation [Fig. 2(c)]. Viability was unchanged by reoxygenation. These resultsindicated that we should use 30 min as the anoxic period to test the role of lactate as a component of either anoxic damage or the oxygen paradox. The damage induced by 60 min anoxia was too great to meaningfully test other components, whereas damage at 30 min anoxia was mild enough that a lactate effect could be detected. The changes shown in Figures 1 and 2 can be prevented by including glucose in the medium. Myocyte preparations are unphysiological in the sense that the cells are not working; glycogen stores are therefore not easily exhausted in thesepopulations. Glucose was kept in the medium and lactate added in order to determine if lactate damaged normal, metabolically competent cells. These data are shown in Figure 3. Figures 3(a), 3(c) and 3(e) illustrate the lack of effect of either the anoxia protocol or lactate on the myocytes when glucosewas present in the medium. Cell morphology, adenine nucleotides, inosine monophosphate, and guanine nucleotides were unchanged after 30 min anoxia. Lactate produced no observable changes in these parameters compared to anoxic controls. In a similar way, lactate did not produce damage or alter high energy phosphatesof cells reoxy-

Damage

in Rat Cardisc

genated for 30 min following the 30 min anoxic period [Figs 3(b), 3(d) and 3(f)]. Omission of glucose from the medium did not render the cells more susceptible to damage with lactate present. The data from these experiments are shown in Figures 4(a) through 4(f). Thirty min anoxia decreased the percentage of rod-shaped cells (71y0 vs. 84% in oxygenated controls; Figure 4(a)), which appeared quantitatively as contracted cells (14% vs. 2% in oxygenated controls). Viability was unchanged by this treatment. The percentage of rod-shaped cells was decreased in lactate-containing flasks over anoxic controls, achieving marginal significance at 50 mM lactate. Although lactate produced an apparent increase in contracted cellscompared to anoxic controls, significance at 17; of the F-distribution could not be establishedwith the number of experiments in this study (n = 6). Both ATP and GTP were reduced 50% in those cellswhich were anoxic for 30 min [Figs 4(c) and 4(e)]. Total adenine nucleotides also were depressed and IMP increased five- to seven-fold during the anoxic period. The presenceof lactate did not affect thesevalues. Reoxygenation of myocytes after 30 min anoxia decreased the percentage of rodshaped cells as compared to aerobic controls, but did not significantly affect viability [Fig. 4(b)]. Lactate in the medium caused no further damage over reoxygenated controls. Nucleotide contents generally returned to preanoxic values with the exception of those cellsin 50 mM lactate, in which ATP remained lower than 60 min aerobic controls (P < 0.05). This result was not significant compared to the reoxygenated controls (P > 0.05). No significant changesdue to lactate could be established. Additional studies were done to determine whether increased proton concentration was necessaryto generate lactate-enhanced reoxygenation damage. This set of experiments, like the last, used zero-glucose medium, but the pH was 6.9 rather than 7.3. The results of these experiments are shown in Figures 5(a) through 5(f). Changes in morphology and viability shown in Figures 5(a) and 5(b) are virtually identical to the pattern demonstrated in Figures 4(a) and 4(b). Nucleotide changes shown in Figures 5(c) through 5(f),

1333

myocytes

although numerically lower than those in Figures 4(c) through 4(f), are identical in pattern to these latter values. Discussion

Lactic acid accumulation in ischemic heart tissuehas been proposed asone of the primary causes of the damage observed in postischemic heart (Neely and Grotyohann, 1984). We tested this hypothesis using a system in which isolated cardiac myocytes were slightly damaged by exposure to anoxia. Lactate was used in two of the four protocols and allowed to equilibrate with the cells. The premise of this approach is that if lactate is truly the key trigger of reoxygenation damage, it should have produced damage additional to that seenin our anoxia/reoxygenation protocols without added lactate. Lactate can enter the myocardial musclecell via a specific translocase (Dennis et al., 1985; Kammermeier et al., 1985; Trosper and Philipson, 1987). Addition of 10 mM or 50 mM lactate should have greatly exceeded any amount present in the cells, thus generating an inwardly-directed gradient. If lactate by itself causesdamage in cardiac muscle, it should have been evident in the isolated cell preparation. Both anoxic and anoxia/reoxygenation damage were demonstrated in the isolated cardiac myocyte preparation. The formation of shortened cells (designated “contracted” in this study) after exposure to anoxia is well documented (Haworth et al., 1981; Hohl et al., 1982; Stern et al., 1985). These cells presumably correspond to the “R-form”, or recontraction phase, described by Stern et al. ( 1985). These “reversibly contracted” cells (cj Briefley et al., 1986) differ from the completely rounded forms in the organization of the myofibrils. Reversibly contracted cells have not been damaged sufficiently to cause that forceful contracture capable of disorganizing the contractile apparatus. “Contracture” a pears to be related to the elevation of free CaF+ and lack of Mg-ATP in the cytosolic compartment (Altschuld et al., 1985). Restoration of Mg-ATP is apparently sufficient to induce hypercontracture in contracted cells. In our preparation, total rod-shaped cells decreased dramatically in favor of the contracted cells after 60 min anoxia, and in favor

1334

T. P. Geisbuhler

of hypercontracted cells after 60 min anoxia and reoxygenation. Viability remained high, however, even under conditions where most of the cells were hypercontracted. Although the reversibility of the damage represented by hypercontracture can be debated, it is clear that these cells are damaged. It is also clear that neither contracted nor hypercontracted cells represented the majority of cells after 30 min anoxia or 30 min anoxia/30 min reoxygenation. At 60 min anoxia, the majority of cells in the flask are damaged. For this reason, 30 min was chosen as the “base period” of anoxia, which we then used to test lactate effects. Lactate did not causeadditional damage to myocytes exposed to anoxia or anoxia/ reoxygenation under most conditions studied. At 50 mM lactate and pH = 7.3, 30 min anoxia significantly reduced the rod-shaped cells as compared to anoxic controls. This significance was marginal, however, and of questionable importance, as both contracted cells and hypercontraccted cells were subtracted from the total cells to obtain this number. In addition, no such significance could be established under identical conditions at pH = 6.9, suggesting that the observed damage was not a reproducible phenomenon. Lactate addition also had no effect on purine nucleotides. By inference, one could say that becauseATP levels were not much affected by lactate under anoxic conditions, the effect of lactate on glycolytic rates in this system were probably minimal. Of particular interest to us was the fact that lower pH did not enhance the ability of lactate to damage the cells. More lactate should have been transported into the cells at lowered external pH, resulting in both increased intracellular lactate and proton concentrations. Not only was cell morphology unaffected by the added lactate, but purine nucleotide levels as well. We concluded from the data that lactate by itself could not be a major participant in the generation of anoxic damage or

and M. J. Rovetto

the oxygen paradox. This conclusion would seemto directly contradict earlier studieswith whole heart (Neely and Grotyohann, 1984). This is not necessarily the case,asour test systemsare different in several important ways. First, the system used in the Neely and Grotyohann study was a working heart preparation. Our quiescent celIs do not perform mechanical work. Second, Neely and Grotyohann useda mechanical function parameter as an index of damage, and measuredlactate in the effluent medium to correlate with this damage. This approach is excellent for correlating function with metabolic phenomena; however, metabolic changeswhich do not result in a product (for example, changesin redox poise)or which produce products which would be masked or buffered by the perfusate (production of proton) could conceivably be undetected in such a system. Our isolated myocyte systemdid not suffer from this handicap, as the bathing medium was used as a tool to manipulate conditions of lactate and pH within the cell. Although great care must be used in comparing suspended or cultured myocytes with intact hearts, the biochemical similarities between the two systemsare sufficient to warrant drawing conclusions regarding lactate in this study. The lack of damage attributable to added lactate is convincing evidence that the accumulation of lactate anion is not the mechanism by which anoxia and reoxygenation damage heart tissue.

Acknowledgements

This study was supported by grants from the National Institutes of Health (HL-27336, HL-36826, and HL-39025). The technical assistanceof Ann Long and Donald Wycoff is gratefully acknowledged. A preliminary account of a portion of this work has been presented [Geisbuhler and Rovetto, FASEB J 4: A419 (No. 880) (1990)].

References RA, WENGER WC, LAMKA KG, KINDIG OR, CAPEN CC, MIZUHIRA V, VANDER HEIDE RS, BRIERLEY GP (1985) Structural and functional properties of adult rat heart myocytes lysed with digitonin. J Biol Chem 260: 14325-14334. BRIERLEY GP, WENCER WC, ALTSCHULD RA (1986) Heart myocytes as models of the cellular response to ischemia. Adv Exp Biol Med 194: 303-314.

ALTSCHULD

Damage

in Rat

Cardiac

myocytes

133.5

SC, KOHN MC, ANDERSON GJ, GARFINKEL DL (1985) Kinetic analysis ofmonocarboxylate uptake into perfused rat hearts. J Mol Cell Cardiol 17: 987-995. GEISBUHLER T, ALTSCHULD RA, TREWYN RW, ANSEL AZ, LAMKA K, BRIERLEY GP (1984) Adeninc nucleotide metabolism and compartmentalization in isolated adult rat heart cells. Circ Res 54: 536-546. GEISBUHLER TP, JOHNSON DA, ROVETTO MJ (1987) Cardiac myocyte guanosine transport and metahoiism. Am .J Physiol 253: C645-C65 1. HAWORTH RA, HUNTER DR, BERKOFF HA (1981) Contractwe in isolated adult rat heart cells. Role of Ca*+. ATP. and compartmentation. Circ Res 49: I1 19-l 128. HOHL CM, ALTSCHULD RA, BRIERLEY GP (1983) Effects of calcium on the permeability of isolated adult rat heart ceils to sodium. Arch Biochem Biophys 221: 197-205. HOHL C, ANSEL A, ALTSCHULD R, BRIERLEY GP (1982) Contracture of isolated rat heart cells on anaerobic to aerobic transition Am J Physiol 242: H1022-H1030. JENNINGS RB, REIMER KA (1981) Lethal myocardial ischemic injury. Am J Pathol 102: 241-255. KAMMERMEIER H, WEIN B, GRAF W (1985) Characteristics of lactate transfer in isolated cardiac myocytes. Basic Res Cardiol 80 [Suppl I): 57-60. LOWRY OH, ROSEBROUGH NJ, FARR AL, RANDALL RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265-275. NEELY JR, GROTYOHANN LW (1984) Role of glycolytic products in damage to ischemic myocardium. Dissociation of adenosine triphosphate levels and recovery of function of reperfused ischemic hearts. Circ Res 55: 816-824. STERN MD, CHIEN AM, CAPOGROZZI MC, PELTO DJ, LAKATTA EG ( 1985) Direct observation of the “oxygen paradox” in single rat ventricular myocytes. Circ Res 56: 899903. TROSPER TL, PHILIPSON KD (1987) Lactate transport by cardiac sarcolemmal vesicles. Am J Physiol 252: C483-C489. WILLIAMSON JR, FORD C, ILLINGWORTH J, SAFER B (1976) Coordination of citric acid cycle activity with electron transport flux. Circ Res 38 [Suppl I]: 139-151. DENNIS