α B Crystallin Translocation and Phosphorylation: Signal Transduction Pathways and Preconditioning in the Isolated Rat Heart

J Mol Cell Cardiol 33, 1659–1671 (2001) doi:10.1006/jmcc.2001.1418, available online at http://www.idealibrary.com on


B Crystallin Translocation and Phosphorylation: Signal Transduction Pathways and Preconditioning in the Isolated Rat Heart ∗Philip Eaton, William Fuller, James R. Bell and Michael J. Shattock The Centre for Cardiovascular Biology and Medicine, King’s College London, The Rayne Institute, St Thomas’ Hospital, London SE1 7EH, UK (Received 11 January 2001, accepted in revised form 31 May 2001) P. E, W. F, J. R. B  M. J. S.
B Crystallin Translocation and Phosphorylation: Signal Transduction Pathways and Preconditioning in the Isolated Rat Heart. Journal of Molecular and Cellular Cardiology (2001) 33, 1659–1671. In this program of studies we have characterized in detail the translocation (assessed by Triton-insolubility) and phosphorylation (using serine-45 or -59 phosphospecific antibodies) of
B crystallin during myocardial ischemia [both with or without ischemic preconditioning (IPC)]. Pharmacological activators and inhibitors allowed us to characterize the signaling pathways involved in
B crystallin phosphorylation during ischemia. Ischemic preconditioning alone caused 30% of the heart’s
B crystallin pool to translocate, providing a significant translocation ‘head-start’ in protected tissue. This enhanced translocation is coupled with increased (3-fold)
B crystallin phosphorylation at both serine residues. The possible role of
B crystallin in the protection afforded by ischemic preconditioning is supported by the signal transduction data; which showed preconditioninginduced
B crystallin phosphorylation can be blocked by tyrosine kinase inhibition (using genistein) and by p38 MAP kinase or PKC inhibition (using SB203580 or bisindolylmaleimide, respectively). The activation of both p38 MAP kinase and PKC are recognized requirements for the induction of preconditioning and their inhibition is known to block protection. Western immunoblotting analysis after isoelectric focusing electrophoresis, confirmed the observations made with the phosphospecific antibodies; but also showed that 27±4% of total cardiac crystallin was phosphorylated after 30 min of ischemia.
B crystallin exists as large polymeric aggregates in cardiac tissue under basal conditions (≈1 MDa as determined by gel filtration chromatography). We induced phosphorylation of
B crystallin during aerobic perfusion by the administration of phenylephrine. However this treatment did not alter the molecular aggregate size of
B crystallin. It appears that
B crystallin molecular aggregate size is not simply regulated by phosphorylation.
B crystallin may have a role to play in the myocardial protection induced by ischemic preconditioning, as both translocation and phosphorylation are both accelerated and enhanced by ischemic preconditioning.  2001 Academic Press K W:
B crystallin; Translocation; Ischemic preconditioning; Heart.

Introduction Many proteins are known to change their cellular localization, or translocate, in response to myocardial ischemia. Proteins which translocate include metabolic enzymes,1 receptors,2 ion and

substrate translocators,3,4 signaling molecules5 and stress proteins.6–8 In many cases the biological rational for these translocations is well understood.
B crystallin, a small stress protein, constitutes as much as 3% of the readily soluble protein in the heart and, in quantitative terms, its translocation

∗ Please address all correspondence to: Philip Eaton, Cardiovascular Research, The Centre for Cardiovascular Biology and Medicine, The Rayne Institute, St Thomas’ Hospital, London SE1 7EH. Telephone 020 7928 9292 (ext. 2749). Fax: 020 7922 8139. E-mail: [email protected]

0022–2828/01/091659+13 $35.00/0

 2001 Academic Press

1660

P. Eaton et al.

during ischemia is therefore very significant.9–11 Despite the large translocation of
B crystallin during ischemia, little is known either about the underlying mechanism or its biological significance. As
B crystallin is a small heat shock protein, it has been postulated that this translocation is part of an adaptive, protective mechanism guarding against a variety of stresses. Overexpression of stress proteins (including
B crystallin) in transfected cell lines or transgenic mice leads to enhanced ischemic tolerance.12–15 Because of this potential role in cardioprotection, it has also been suggested that
B crystallin may play a role in ischemic preconditioning (IPC).16 Ischemic preconditioning is thought to involve induction of protection via the phosphorylation of an effector protein. As stress proteins, including
B crystallin, are functionally regulated by phosphorylation, this is consistent with their possible involvement in the ischemic preconditioning process.17–21 De novo synthesis of stress proteins has clearly been implicated in the molecular basis of the second window of ischemic preconditioning.22–24 Thus, it is possible that phosphorylation and translocation of constitutive
B crystallin could contribute to the first window or acute phase of preconditioning. Previously we have shown that IPC causes translocation of
B crystallin phosphorylated at serine-59.16 In this study we have fully characterized the translocation and phosphorylation (at serine-45 and -59) of
B crystallin during myocardial ischemia and ascertained how IPC influences these processes. We have used conventional pharmacological methods to assess the signal transduction events involved in
B crystallin phosphorylation and isoelectric focusing to determine the percentage of
B crystallin which becomes phosphorylated in response to an ischemic intervention. These studies have allowed us to determine if
B crystallin phosphorylation and translocation are independent processes or whether they are mutually coordinated.
B crystallin, like other stress proteins, self-associates to form large multimeric complexes. In many cases phosphorylation, induced by stress or pharmacological kinase activation, is the cue for these complexes to breakdown and this may be crucial for translocation and protection.25,26 We have used gel filtration liquid chromatography to measure the multimeric size of
B crystallin in myocardial tissue and assessed whether this process is regulated by phosphorylation. The integrated approach used in this study has allowed us to build up a detailed picture of how
B crystallin responds to myocardial ischemia and how these events are modulated by IPC.

Materials and Methods Animals Male Wistar rats (200–250 g) were used throughout this study and were obtained from BK Universal. The animals were maintained humanely in compliance with the ‘‘Principles of Laboratory Animal Care’’ formulated by the National Society for Medical Research and ‘‘Guide for Care and Use of Laboratory Animals’’ prepared by the National Academy of Sciences and published by the National Institute of Health (NIH publication No. 85–23, revised 1985).

Isolated heart preparation Rats were anesthetised with diethyl ether and injected with sodium heparin (200 IU) via the femoral vein. Hearts were rapidly excised, placed in cold (4°C) bicarbonate buffer and the aorta cannulated. The hearts were then perfused with oxygenated (95% O2+5% CO2) bicarbonate buffer at 37.0°C (pH 7.4). Perfusion was in the non-recirculating Langendorff mode at a constant pressure equivalent to 1000 mm of H2O. The bicarbonate buffer contained (in mmol/l) 118.5 NaCl, 4.7 KCl, 1.18 KH2PO4, 25.0 NaHCO3, 1.2 MgCl2, 1.4 CaCl2 and 11.1 glucose. In some hearts, ventricular function was assessed via an intraventricular balloon inflated to give an initial left-ventricular end-diastolic pressure (LVEDP) of 4 mmHg. Left-ventricular pressure (LVP) was measured throughout the protocol and recovery of developed pressure (LVDP) after 60 min of reperfusion was expressed as a percentage of the initial value measured at the end of the stabilization period.

Perfusion protocols The protocols used in this program of studies are shown sequentially in Figures 1, 2, 3, 4 and 6 (Panels A), along with the data obtained using them. In studies involving whole heart zero-flow ischemia, this was initiated by terminating perfusion, during which time hearts continued to be maintained at 37°C in a thermostatically-controlled chamber. Kinase activators and inhibitors were purchased from Calbiochem-Novabiochem Ltd (Nottingham, UK) and used in accordance with the suppliers instructions. Genistein, SB203580, bisindolylmaleimide and PD98059 were used at


B Crystallin Translocation and Phosphorylation

concentrations of 50 , 10 , 10  and 10  respectively. Anisomycin, phenylephrine and PMA were used at a concentrations of 50 ng/ml, 10  and 200 n respectively.

Protein analysis Tissue was homogenized in 20 m Tris-HCl, pH 7.4+phosphatase inhibitors NaF and Na3VO4 at 5 m and 1 m respectively (10 ml of buffer/g wet wt of tissue), using a Polytron homogenize. When detergent soluble and insoluble compartments were required, 0.1% Triton ×100 was added to the homogenization buffer, prior to centrifugal fractionation at 20 000 g for 5 min. After homogenization and fractionation (if required) proteins were then reconstituted in SDS sample buffer (Laemelli, 1970). SDS-polyacrylamide gel electrophoresis (PAGE) was carried out using the BioRad mini Protean II system on 12.5% polyacrylamide gel. After electrophoresis, samples were transferred to PVDF membrane (Amersham, UK) using a Pharmacia semi-dry blotter. Rabbit polyclonal antibodies to
B-crystallin or HSP27 (Bioquote Ltd., York, UK) were used to probe for the presence of these proteins. Phosphorylated
B-crystallin was detected using phospho-specific antibodies generated in rabbits (generously provided and characterized by Dr Kanefusa Kato).17,27 Rabbit polyclonal antibodies to dual phosphorylated ERK1/2 or p38 MAP Kinase were used to assess kinase activation (New England Biolabs Inc., Herts, UK) An anti-rabbit secondary antibody, coupled to horseradish peroxidase, (Amersham Pharmacia Biotech UK Ltd., Bucks, UK) was used with the enhanced chemiluminescent (ECL) reagent (Amersham) to visualize primary antibody binding. Western blots were digitized using a flat-bed scanner (HP Scanjet 11C) and quantitatively analyzed using the NIH-Image software (Freeware, NIH, Baltimore, USA).

Isoelectric focusing gel analysis Samples for IEF analysis were reconstituted in loading buffer which contained 2.85 g urea, 0.1 ml Triton ×100, 0.1 ml, NP40, 0.1 g CHAPS, 250 l 40% ampholytes (4 parts pH 6–8, 1 part 3.5–10), 0.25 ml -mercaptoethanol, 50 l 10% APS, 5 l TEMED and water (to a final volume of 5 ml). IEF gels were cast using the BioRad mini Protean II system and comprised 2.76 g urea, 0.1 ml Triton ×100, 0.1 ml, NP40, 0.1 g CHAPS, 0.67 ml 30% acrylamide, 0.38 ml ampholytes (4 parts pH 6–8,

1661

1 part 3.5–10) and water (to a final volume of 5 ml). The Anode buffer (lower reservoir) consisted of 2.47 ml 85% phosphoric acid in 500 ml water. The Cathode buffer (upper reservoir) consisted of 7.25 g lysine and 8.71 g arginine in 500 ml water. Proteins were focused overnight (14–16 h) at 220 V and then transferred to PVDF with a Pharmacia semi-dry blotter (100 mA per mini gel for 2.5 h) using a buffer that consisted of 20 ml methanol and 0.7 ml glacial acetic acid in 100 ml water.

Gel filtration 100l Triton ×100 soluble protein (prepared as described above) was injected on to a Sephracyl gel filtration column (Amersham Pharmacia, UK) conditioned with phosphate buffered saline [pH 7.4 (PBS)] pumped at a flow rate of 1 ml/min via a Shimadzu chromatograph utilizing diode array detection. Three-milliliter fractions were collected for 3 h by the automated collector and 15 l of each fraction was dot-blotted to nitrocellulose membrane (Amersham). This membrane was then treated in the same manner as a Western blot and probed with an antibody to
B crystallin to ascertain which molecular weight fractions the crystallin eluted in. A calibration line (shown in Figure 8) was constructed using thyroglobulin (669 KDa), catalase (232 KDa), albumin (67 KDa), chymotrypsinogen (25 KDa) and ribonuclease A (13.7 KDa) as described in the Amersham Pharmacia technical notes. The bed volume of the column was 220 ml and the void volume was determined to be 38 ml using unretained blue dextran.

Statistics Results are presented as mean±... (n=4 per group). Differences between groups were assessed using ANOVA, followed by a Dunnett Test. Differences were considered significant at the 95% confidence level.

Results Preconditioning confers protection against ischemia Preconditioning (3 cycles of 5 min ischemia and 5 min reperfusion) conferred significant protection against post-ischemic contractile dysfunction following a subsequent 35 min index ischemia. Hearts

1662

P. Eaton et al.

A

A Control ischemia alone

Control ischemia 10 min

10 min

30 min

40 min

Preconditioning + ischemia

Preconditioning + ischemia 5'

10 min

5'

5'

5'

5'

5'

10 min

5'

5'

5'

5'

5'

5'

40 min

30 min

B

B

PSer-45 PSer-59

* 0 1 2 3 5 15 20 30 40 5 15 20 30 40 preconditioning mins of ischemia after mins of ischemia cycles ischemic preconditioning

75

50

* *

25

Control ischemia

600

*

300 Control ischemia

0

0

0

30 5 Ischemic duration (min)

Figure 1 (A) Perfusion protocols used to assess
B crystallin translocation. indicates time points where hearts were prepared for analysis. (B) Graph showing IPC enhances
B crystallin translocation during ischemia. ∗ indicates a significant difference between the control and preconditioned groups.

reperfused for 60 mins after the control and preconditioning protocols shown in Figure 1A recovered 46±4% (n=12) and 64±3% (n=10) (P<0.05) respectively. Preconditioning enhances
B crystallin translocation Figure 1B shows the linear translocation of
B crystallin to the insoluble fraction during 30 min of myocardial ischemia. Three cycles of IPC alone were sufficient to translocate significant amounts (30±8%, P<0.05) of
B crystallin. This translocation during IPC has the effect of significantly (P<0.05) increasing the amount of translocated
B crystallin at any given time in the subsequent ischemic period. Thus, after 5 min and 30 min of global ischemia there was significantly more translocated
B crystallin (648% and 160% respectively) in the preconditioned heart than in the non-preconditioned controls.

Preconditioning * + ischemia

0 10 20 30 40 Ischemic duration (min)

αB crystallin serine-59 phosphorylation

C

Preconditioning + ischemia

αB crystallin serine-45 phosphorylation

% translocation of total αB crystallin

100

1000

Preconditioning * + ischemia * *

*

*

500 Control ischemia

0

0 10 20 30 40 Ischemic duration (min)

Figure 2 (A) Perfusion protocols used to assess
B crystallin phosphorylation. indicates time points where hearts were analyzed for total and phosphorylated
B crystallin by Western immunoblot. (B) Western immunoblots showing the phosphorylation of
B crystallin at serine-45 and -59 following IPC or ischemia. (C) Quantitation of Western immunoblots, comparing
B crystallin phosphorylation (at serine-45 and -59) during ischemia in control and preconditioned hearts. ∗ indicates a significant difference between the control and preconditioned groups.

Preconditioning enhances
B crystallin phosphorylation Figures 2B and C show the time-course of
B crystallin phosphorylation during ischemia (with or without IPC) at serine-45 and -59. Preconditioning induced significant (P<0.05) phosphorylation of
B crystallin at both serine residues prior to the onset of ischemia. This prephosphorylation of
B crystallin resulted in preconditioned hearts having increased levels of phosphocrystallin during a subsequent ischemic episode, compared with non-preconditioned controls. This enhanced level of phosphocrystallin in preconditioned tissue is particularly notable during the first 10–20 min of ischemia; but become less marked as ischemia is extended to 40 min. It is of note that none of the experimental interventions used in these studies caused phosphorylation of
B crystallin at serine-19 (data not shown).


B Crystallin Translocation and Phosphorylation A Aerobic perfusion 10 min agonist

Preconditioning cycles 10 min

5 min

5 min

5 min

5 min

5 min

5 min

Serine-45 phosphorylation

B 600 * 400

*

Control

Serine-59 phosphorylation

IPC

Anisomycin PMA Phenylephrine

C 400

*

*

300

* *

200 100 0

Control IPC

aerobic perfusion. However, following administration of phenylephrine to the aerobicallyperfused isolated heart enhanced serine-45 and -59 phosphorylation occurs (Fig 3B/C). Despite this marked phenylephrine-induced phosphorylation we detected no translocation of total
B crystallin (4.3±2.1%), phosphoserine-45
B crystallin (1.2±0.2%) or phosphoserine-59
B crystallin (2.51.2%). This contrasts 30 min of ischemia which was less effective in inducing phosphorylation than phenylephrine treatment, but efficiently translocated 57.2±7.1% of total
B crystallin.
B crystallin phosphorylation – signal transduction

200

0

1663

PMA Anisomycin Phenylephrine

Figure 3 (A) Perfusion protocols used in to assess
B crystallin phosphorylation following pharmacological kinase activation for 5 min during aerobic perfusion. indicates where hearts were analyzed for phosphorylated
B crystallin (B, C) Quantitative Western immunoblot data showing the ability of pharmacological kinase activation and IPC to induce
B crystallin phosphorylation at serine-45 and -59. ∗ indicates a significant difference between control and preconditioned (or drug-treated) groups.


B crystallin phosphorylation and translocation Figure 3B and C show the ability of signal transduction agonists to induce
B crystallin phosphorylation at serine-45 and -59 respectively. All the agonists tested were capable of inducing phosphorylation at both serine residues, except anisomycin which only stimulated serine-59 phosphorylation. We assessed whether the phosphorylation and translocation of
B crystallin are mutually coordinated. Quantitative Western immunoblotting has shown that neither translocation nor phosphorylation of
B crystallin occurs during

Figure 4B and C show the ability of kinase inhibitors to attenuate preconditioning-induced
B crystallin serine-45 and -59 phosphorylation respectively. Tyrosine kinase blockade with genistein, PKC blockade with bisindolylmaleimide or p38 MAP kinase blockade with SB203580 were all effective in preventing preconditioning-induced phosphorylation of
B crystallin serine-45 and-59. In contrast, ERK blockade via MEK inhibition with the inhibitor PD98059 was effective in attenuating the phosphorylation of
B crystallin serine-45; but did not attenuate the preconditioning-induced serine-59 phosphorylation. The data presented in Figure 4C (pharmacological blockade of preconditioning-induced
B crystallin serine-59 phosphorylation) was shown in a previous study,16 and is redrawn and included for completeness. Correlating
B crystallin phosphorylation and kinase activation Figure 5 (Western immunoblots) and Figure 6 (quantitated Western immunoblots) compare
B crystallin serine-45 and -59 phosphorylation profiles with the activation (dual phosphorylation) of p38 MAP kinase and ERK1/2 kinases. It is clear that ERK1/2 activity decreases during ischemia whereas p38 MAP kinase activity increases up to 20 min of ischemia, after which it begins to decline – such that it is back to near control levels by 40 min of ischemia. There is a trend for IPC to increase the rate of the ischemia-induced ERK1/2 inactivation. There is also a trend for preconditioning to potentiate the ischemia-induced activation of p38 MAP kinase, particularly during the early phase of ischemia. It is interesting that IPC has differential effects on the activation/inactivation of these two MAP kinases during ischemia, but the

1664

P. Eaton et al.

A Preconditioning cycles 10 min

5 min

5 min

5 min

5 min

5 min

5 min

Preconditioning cycles in presence of kinase inhibitors 10 min

5 min

5 min

5 min

5 min

5 min

5 min

Serine-45 phosphorylation

Kinase inhibitor

400

*

0

* * Control IPC

Serine-59 phosphorylation


B crystallin multimeric aggregate size

B

200

in Figure 2 which shows ischemia-induced phosphorylation using the phosphospecific antibodies. Quantitation of the IEF Western immunoblots shows that phosphorylation increases with ischemic duration, with 27±4% of the total
B crystallin phosphorylated after 30 min ischemia.

*

Genistein SB 203580 Bisindoly PD 98059

Figure 8 shows the molecular aggregate size of
B crystallin as determined using gel filtration liquid chromatography, as well as the molecular weight standards used to calibrate the Sephracyl column used. Detergent soluble
B crystallin prepared from control or phenylephrine-treated hearts showed no differences in multimeric complex size; and this is despite the fact that phenylephrine induced efficient phosphorylation of
B crystallin. Although it is of note that both preconditioning and anisomycin treatment were more effective at phosphorylating
B crystallin at serine-45 than phenylephrine.

C 400

Discussion

200

0

*

* *

Control IPC

Genistein SB 203580 Bisindoly PD 98059

Figure 4 (A) Perfusion protocols used to assess
B crystallin phosphorylation in IPC hearts with or without pharmacological blockade of kinase pathways. indicates where hearts were analyzed for phosphorylated
B crystallin (B, C) Quantitative Western immunoblot data showing preconditioned-induced phosphorylation of
B crystallin (at serine-45 and -59); and the ability of various kinase inhibitors to attenuate this phosphorylation. ∗ indicates a significant difference between preconditioned and preconditioned plus kinase inhibitor groups.

differences between the ischemic and conditioned-ischemic profiles are minimal.

pre-

How much
B crystallin is phosphorylated? Figure 7 shows IEF Western immunoblots probed with an antibody that recognizes total crystallin. It is clear that ischemia results in the generation of crystallin with a lower pI, which is interpreted as phosphorylated crystallin, in light of the data shown

The translocation of
B crystallin from a detergentsoluble to a detergent-insoluble fraction during myocardial ischemia is a robust observation,1,6–8 but has not been rigorously investigated.
B crystallin is also a substrate for kinases which become activated during ischemia/preconditioning, but little is known about the regulation of the phosphorylation status of this protein in cardiac tissue. In these experiments we have fully characterized both
B crystallin translocation and phosphorylation during a detailed time-course of myocardial ischemia. As
B crystallin is a small heat-shock or stress protein with the ability to confer tolerance to cellular stresses, such as ischemia,12–14 it has been suggested that this protein may have a role to play in the cardioprotective phenomenon of IPC. Consequently, we have extended our study to establish if IPC modulates the time-course of
B crystallin translocation and phosphorylation during ischemia and to characterize the kinases involved.

Ischemia induces translocation and phosphorylation of
B crystallin The linear ischemia-induced translocation observed in this study is in agreement with previous reports from studies in rat, guinea pig and human tissue; as well as those studies which simulated ischemia


B Crystallin Translocation and Phosphorylation

1665

Ser-45 crystallin phosporylation Ser-59 crystallin phosporylation

p38 activation

ERK 1/2 activation 0

1

2

Preconditioning cycles

3

5

15

20

30

40

Min of ischemia after preconditioning

5

15

20

30

40

Min of ischemia

Figure 5 Western immunoblots showing the change in
B crystallin serine-45 phosphorylation,
B crystallin serine59 phosphorylation, p38 MAP kinase dual phosphorylation (activation) and ERK1/2 dual phosphorylation (activation) during ischemia (in the presence or absence of preconditioning). Quantitation of these immunoblots is shown in Figure 6.

in isolated cell models.1,6–8,28 Additionally, we have demonstrated that increased phosphorylation of
B crystallin occurs at both serine-45 and -59 as ischemia proceeds.

Preconditioning cycles enhance
B crystallin translocation and phosphorylation It is clear that IPC, prior to the onset of an index ischemia, induced substantial
B crystallin translocation (amounting to almost one third of the total pool). As
B crystallin remains associated with the Triton-insoluble compartment for up to two hours after translocation,16 when a preconditioned heart enters the index ischemia it has a substantial ‘headstart’ in terms of translocated
B crystallin. Not only is
B crystallin translocation influenced by preconditioning but also the phosphorylation state of the protein is altered after preconditioning and before the index ischemia. Consequently, as the preconditioned heart enters a long period of ischemia, a significant fraction of the
B crystallin pool is already phosphorylated. Furthermore, this increased phosphorylation is maintained and even increased throughout the ensuing index ischemia. In control hearts, although the phosphorylation status of
B crystallin is low at the start of the index ischemia, the rate of phosphorylation is faster than in the preconditioned hearts. Thus, by 40 min of ischemia, there is no difference in the phosphocrystallin status of control and preconditioned hearts. This accelerated rate of phosphorylation in control ischemia may reflect regulation of
B crystallin phosphorylation status and a mechanism

which tightly governs the degree to which this protein is phosphorylated. Our observations regarding differential translocation and phosphorylation contrast with those of Armstrong et al.,29 who were unable to detect differential translocation or phosphorylation of
B crystallin in rabbit cardiomyocytes preconditioned with simulated ischemia. This is likely to reflect differences between the two models used to investigate the role of
B crystallin in IPC. If
B crystallin translocation has any functional consequence in the ischemic heart, clearly the preconditioned myocardium has a ‘head-start’ with this process. Similarly, compared to control, the preconditioned heart enters ischemia with an elevated phosphocrystallin status. Preconditioning has the effect of shifting the temporal profile of
B crystallin phosphorylation during ischemia to the left. This means the preconditioned heart obtains an
B crystallin phosphorylation status it would take a control heart up to 20 min to achieve. If the functional role of
B crystallin translocation and phosphorylation is to induce myocardial protection or adaptation to ischemia, the fact that these events are both brought forward and enhanced could explain the protection observed by preconditioning. The temporal shift in phosphorylation and translocation correlates approximately with the delay in ischemic injury induced by preconditioning. Although consistent with the hypothesis that phosphorylation and translocation of small heat shock proteins may play a role in the protection induced in the acute phase of preconditioning, further studies are necessary to prove a causal relationship. If the translocation, phosphorylation and dis-

1666

P. Eaton et al.

Ser-45 phosphorylation

A *

600 *

*

*

*

*

* *

*

*

*

*

*

Ser-45 crystallin phosphorylation

0

Ser-59 phosphorylation

B *

1200 *

*

*

*

*

* *

*

Ser-59 crystallin phosphorylation

*

*

0

Dual phosphorylation

C *

750

* *

*

*

*

p38 activation

* 0

Dual phosphorylation

D 200 ERK 1/2 activation *

0

1 2 3 Control Preconditioning cycles

* * 15 20 30 40 Min of ischemia after preconditioning 5

* 5

15 20 30 40 Min of ischemia

Figure 6 (A, B)
B crystallin serine-45 and -59 phosphorylation profiles during ischemia (in the presence or absence of preconditioning). (C, D). p38 MAP Kinase and ERK1/2 Kinase activation (dual phosphorylation) profiles during ischemia (in the presence or absence of preconditioning). ERK1/2 activity decreases during ischemia whereas p38 MAP kinase activity increases during the first 20 min of ischemia, after which it begins to decline to basal levels. There is a trend for IPC to increase the rate of the ischemia-induced ERK1/2 inactivation. There is also a trend for preconditioning to potentiate the ischemia-induced activation of p38 MAP kinase, particularly during the early phase of ischemia. The p38 MAP Kinase (but not the ERK1/2 Kinase) activation profile could account for the
B crystallin serine-45 and -59 phosphorylation profile. Cycles of IPC result in
B crystallin serine-45 and -59 phosphorylation. This has the effect of enhancing
B crystallin serine-45 and -59 phosphorylation in the preconditioned ischemic heart compared to control ischemic preparations – and effect which is particularly notable during the first 20 min of ischemia. ∗ indicates a significant difference from the control group.

aggregation of small heat shock protein do play a role in the myocardial response to ischemia, then the question remains as to what, on a molecular level, these proteins actually do? Suggested cellular roles for stress proteins include molecular chaperone activity, the strengthening of the cytoskeleton and protection of proteins at risk of proteolysis following calcium overload.26,30–33 Of these possibilities, the strengthening of the cytoskeleton may be particularly important as Jennings and colleagues34,35 and others36 have shown that a principal mechanism of damage during ischemia and reperfusion is cell swelling. Preconditioning could thus reduce swelling-induced damage by either decreasing the propensity of myocytes to

swell (by osmotic unloading them) or by strengthening the cytoskeleton to withstand better the osmotic stress itself. Preconditioning may be predicted to decrease molecular chaperone activity, as phosphorylation of chaperone proteins leads to breakdown of their functional aggregate, attenuating their chaperone activity.26,30 Signal transduction and
B crystallin phosphorylation The profile of p38 MAP kinase activation during preconditioning and ischemia could explain the phosphorylation profile for
B crystallin. In contrast it is unlikely that ERK activation can directly explain


B Crystallin Translocation and Phosphorylation A Ischemia time course 30 min

B Crystallin Phosphocrystallin Phosphocrystallin 0

7.5 15 22.5 30 Min of ischemia

Decreasing pI

20 min

Figure 7 (A) Perfusion protocols used in to assess
B crystallin phosphorylation by isoelectric focusing electrophoresis/Western immunoblotting. indicates time points where hearts were prepared for analysis. (B) Western immunoblot probed with an
B crystallin antibody, showing the generation of phosphocrystallin as ischemia proceeds. Following 30 min of ischemia, 27% of total cardiac crystallin is phosphorylated.

= aerobic control: 1.06 +/– 0.063 MDa X = phenylephrine: 0.97 +/– 0.045 MDa

0.4

0.3

1

K (av)

2

0.2 y = –0.131LOG(x) + 0.834

3 4

0.1

5. Thyroglobin (669 KDa) 5 4. Catalase (232 KDa) 3. Albumin (67 KDa) 2. Chymotrypsinogen (25 KDa) 1. Ribonuclease A (13.7 KDa)

0 1000

X

10 000 100 000 1 000 000 10 000 000 log MWT protein

Figure 8 Graph showing the calibration of the gel filtration column used to assess the molecular aggregate size of
B crystallin. There is no significant difference in the molecular aggregate size of
B crystallin from aerobic control (Α) or phenylephrine-treated (×) tissue.


B crystallin phosphorylation, as ERK activation decreases as ischemia proceeds and this happens at the same time that crystallin becomes phosphorylated. There appears to be a trend for preconditioning to speed up the rate of ischemiainduced ERK inactivation.
B crystallin phosphorylation by MAPKAP kinase 2 may be predicted as it is this kinase which phosphorylates its sister protein HSP27.37 However, in contrast to HSP27,

1667

the evidence that MAPKAP kinase 2 phosphorylates
B crystallin in vivo is not conclusive; although there has been a recent report in support of this possibility.38 Tentative evidence that MAPKAP kinase 2 phosphorylates
B crystallin can be taken from this study, as preconditioning-induced phosphorylation of
B crystallin could be blocked efficiently by SB203580, the potent inhibitor of p38 MAP kinase which is directly upstream of MAPKAP kinase 2. In our studies, p38 MAP kinase activation by anisomycin appears to selectively phosphorylate serine-59 but not serine-45. This is consistent with the observations of others who reported
B crystallin phosphorylation at serine-59 is via p38 MAP kinase while that at serine-45 is via the ERK pathway.27,38 The role of the ERK pathway in serine-45 is confirmed in our studies since PD 98059 (a specific inhibitor of the ERK pathway) blocks preconditioning-induced phosphorylation at serine-45 but not at -59. However, it is difficult to simply attribute serine-45 phosphorylation to ERK and serine-59 phosphorylation to p38 MAP kinase, as PMA (which is known to activate ERK) also induced serine-59 phosphorylation. This discrepancy is further highlighted by the observation that PKC inhibition with bisindolylmaleimide attenuated both serine-45 and serine-59 phosphorylation. Theoretically, it might be expected that serine-59 would be phosphorylated via p38 MAP kinase, as phosphorylation by MAPKAP kinase 2 occurs at an Arg-X-X-Ser motif, as observed at the three HSP27 phosphorylation sites. Similarly, the serine-45 phosphorylation site is within the ERK phosphorylation motif Leu-Ser-Pro. However, in the isolated rat heart complex signal transduction regulates
B crystallin phosphorylation and it is not possible to conclude from our data that a specific kinase is responsible for the phosphorylation of a specific phosphorylation site. To attribute a specific protective effect to the phosphorylation of serine-59 would be misleading as phenylephrine has been shown to be a potent initiator of preconditioning in the rat heart39 while in the present study it induces phosphorylation most effectively at serine-45. Similarly, while anisomycin appears to selectively phosphorylate serine-59, preconditioning itself phosphorylates both serine residues equally – with this preconditioning-induced phosphorylation of both sites being inhibited by SB203580. The explanation for these apparently contradictory effects is unclear but these data may indicate, as discussed above, that different kinase pathways phosphorylate specific residues on
B crystallin and, as previously reported,40 SB 203580 is not specific for p38 MAP kinase.

1668

P. Eaton et al.

The activation of PKC has also been repeatedly implicated in the pathway leading to preconditioning-induced protection.41 Our data suggest that preconditioning-induced phosphorylation of both serine-59 and serine-45 is inhibited by the PKC inhibitor bisindolylmaleimide. However, the effects of stimulating PKC with PMA were equivocal. Thus, although there is clear cross-talk and overlap between kinase pathways, both PKC and p38 MAP kinase are able to phosphorylate
B crystallin. The involvement of these kinases, previously implicated in preconditioning-induced protection, reinforces the concept that the phosphorylation of small heat shock proteins could play a role in mediating cardioprotection. IPC has a differential effect on the activation/ inactivation of these two MAP kinases during ischemia. This trend for preconditioned ischemic hearts to show potentiated p38 MAP kinase activation during ischemia, and at the same time attenuated ERK 1/2 inactivation, is interesting. However, the absolute differences between these two groups are minimal and thus preclude discussion of how this differential regulation of these kinases during ischemia could be protective.

we have used aerobic, acidic (pH 6.8) bicarbonate buffer perfusion to elicit translocation and we were able to show that the translocated
B crystallin was not phosphorylated.16 Thus, phosphorylation does not appear to be either necessary or sufficient to induce translocation which may be separately regulated by other factors such as intracellular pH.

How much crystallin becomes phosphorylated during ischemia? It is clear that ischemia induces crystallin phosphorylation. However, as this observation has been made using phosphospecific antibodies, this technique gives no indication of how much of the total cellular pool of crystalline becomes phosphorylated. To address this issue, we used isoelectric focusing to compare the phosphorylation status of
B crystallin in control and ischemic tissue. It is apparent that most of the
B crystallin remains unphosphorylated even after 30 min of ischemia. This is in agreement with a previous report, in a different model, showing heat stress induces phosphorylation of only a small fraction of the total cardiac
B crystallin.17

Is phosphorylation and translocation of
B crystallin linked?

Small stress protein phosphorylation and aggregate size

In these studies, whether in control or preconditioned ischemia, we observed a temporal association between
B crystallin phosphorylation and its translocation. This raises the possibility that the two processes may be causally related. For example, it is possible that phosphorylation of
B crystallin is an essential trigger for its translocation. To address this possibility, we have compared the amount of translocated
B crystallin in aerobic control hearts with those treated with phenylephrine. Although phenylephrine induced efficient phosphorylation at serine-45 and -59, it did not enhance translocation of either total or phosphorylated
B crystallin. It might therefore be concluded that phosphorylation of
B crystallin is not requisite for its translocation.
B crystallin translocation can be achieved in vitro by simply acidifying cardiac homogenates or by perfusing isolated hearts with mildly acidic (pH 6.8) bicarbonate buffers.1,16,42 During ischemia
B crystallin phosphorylation occurs at the same time as the cellular pH drops; so if these processes are dual requirements for translocation this could explain why phenylephrine treatment alone during aerobic perfusion did not induce translocation. Previously,

Small stress proteins like
B crystallin and its sister protein HSP27 exist in monomeric and polymeric forms. In general, HSP27 is unphosphorylated under stress-free condition and exists as large polymeric aggregates. Activation of MAPKAPK2 via p38 MAP kinase during stress leads to HSP27 phosphorylation and breakdown of the polymeric aggregate.26,43 The introduction of the negative charge by phosphorylation seems to be crucial in mediating breakdown of the complex and has been confirmed using site-directed mutation experiments where the introduction of an acidic amino acid at the phosphorylation site leads to the dominance of monomeric HSP27.25
B crystallin and HSP27 show significant homology and thus it might be expected that regulation of
B crystallin aggregate size would also involve phosphorylation. In this final part of our study, we have assessed the aggregate size of
B crystallin using gel filtration and found that under control conditions it has a molecular mass approximately 1 MDa. Phenylephrine efficiently induced phosphorylation of serine-45, but did not result in breakdown of the
B crystallin aggregate; thus
B crystallin aggregate size appears not to be regulated in the same way


B Crystallin Translocation and Phosphorylation

as HSP27. It should however be noted that the phenylephrine-induced phosphorylation of serine45 was not actually as efficient as that stimulated by IPC or anisomycin treatment. It has been reported that negative charges in the C-terminus of
B crystallin are important in the stabilization of the aggregate complex;44 whereas the serine-45 an59 phosphorylation sites within
B crystallin are N-terminal. Regulation of
B crystallin aggregate size by C-terminal charge modification would be consistent with the present observations. It is noteworthy that
B crystallin can be modified at the Cterminal threonine-170 by N-acetylglucosamine, a modification which has been reported to be reduced in the ischemic rat heart.45 It is possible that this modification plays a role in the cellular control of
B crystallin aggregate size. The presence of by Nacetylglucosamine at threonine-170 reduces the pI of crystallin as determined by isoelectric focusing gel electrophoresis analysis.46 This modification at the C-terminus, which introduces a negative charge, may stabilize the crystallin complex. Likewise removal of N-acetylglucosamine from crystallin, which occurs during ischemia, may be predicted to destabilize the complex. Martin et al.,47 have reported that transfection of a mutant
B crystallin, which was unable to form full-size oligomers and did not enhance ischemic tolerance. This could be interpreted that oligomer breakdown has no role in cardioprotection. However, these studies did not actually assess changes in preconditioning and the extent to which the mutant decreased the basal oligomeric size was not reported.

Summary and Conclusions In summary, we have characterized the effect of ischemia and IPC on the translocation and kinasemediated phosphorylation of the small heat shock protein
B crystallin. Short periods of preconditioning ischemia both accelerate ischemiainduced phosphorylation of
B crystallin (at serines -45 and -59) and the translocation of the protein to the detergent insoluble cytoskeletal/myofilament fraction of tissue homogenates. Phosphorylation of
B crystallin at serine-59 appears to be mediated via activation of p38 MAP kinase and preconditioning-induced phosphorylation can be prevented by inhibitors of either PKC or p38 MAP kinase. While preconditioning accelerates ischemiainduced translocation and phosphorylation of
B crystallin, these two process appear to be independent of each other. Phosphorylation is neither sufficient nor necessary to initiate translocation

1669

which may be triggered by other factors such as acidosis. Finally, phosphorylation does not alter the mean aggregate size of the
B crystallin multimeric complex. These data are consistent with the hypothesis that the phosphorylation and translocation of small heat shock proteins to the cytoskeleton/ myofilaments may play a cardioprotective role in IPC. The mechanism by which this is mediated remains to be elucidated, although possibilities include modulation of chaperone function, strengthening of the cytoskeleton and protection of proteins at risk of proteolysis following calcium overload.

Acknowledgements This work was supported by grants from the Wellcome Trust and the British Heart Foundation. We thank Dr Kanefusa Kato (Department of Biochemistry, Institute for Developmental Research, Aichi Human Service Centre, Kasugai, Aichi 4800392, Japan) for kindly providing the antibodies to phosphorylated
B-crystallin.

References 1. B R, M R, D P, C E, S S, DL F. Binding of cytosolic proteins to myofibrils in ischemic rat hearts. Circ Res 1996; 78: 821–828. 2. M AS, M HJ, Z MG, I PA. Ischemia-induced and agonist-induced changes in alpha adrenergic and beta adrenergic receptor traffic in guinea pig heart. Am J Physiol 1987; 253(5): H1159–H1166. 3. E S, N N, S M. Contribution of alpha-adrenergic and beta-adrenergic stimulation to ischemia-induced glucose transporter (GLUT)4 and GLUT1 translocation in the isolated perfused rat heart. Circ Res 2000; 84(12): 1407–1415. 4. S MJ, E P, B G, L JM. Ischemiainduced translocation of the Na/K pump away from the sarcolemma in the isolated rat heart. Circulation 1997; 96: 311. 5. M Y, H T, Y K. Nuclear translocation of PKC zeta druing ischemia and its inhibition by wortmanin, an inhibitor of phosphatidylinositol 2-kinase. FEBS Letters 1997; 401(2–3): 247–251. 6. B F, W A, C M. Alpha B crystallin in cardiac tissue – association with actin and desmin filaments. Circ Res 1992; 71: 288–294. 7. K FAJM, G MLJ, I PRLA, S LHEH, J WW. Alpha B crystallin and HSP25 in neonatal cardiac cells – differences in cellular localization under stress conditions. Eur J Cell Biol 1998; 75: 38–45. 8. C M, B F. Determination of alpha B crystallin aggregation – a new alternative method

1670

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

P. Eaton et al.

to assess ischemic damage of the heart. Basic Res Cardiol 1992; 87: 38–46. K K, S H, K N, I Y, S K, O K. Tissue distribution and developmental profiles of immunoreactive alpha B crystallin in the rat determined with a sensitive immunoassay system. Biochim Biophys Acta 1991; 1074: 201–208. L G, V R, O U, W M, G HJ, K R, B R. Abundance and location of the small heat shock proteins HSP25 and alpha B crystallin in rat and human heart. Circulation 1997; 96: 3466–3476. L G, V R, O U, W M, S I, S J, B R. Location of the small heat shock proteins HSP25 and alpha B crystallin in rat and human heart. Eur J Cell Biol 1997; 74: 43. R PS, C GA, D DK, E RM, D WH. Genetically engineered mice overexpressing alpha beta crystallin exhibit protection against myocardial ischemia-reperfusion injury. J Am Coll Cardiol 2000; 35(2): 401A. (Abstract). B WF, M JL, M R, D WH. Specific heat shock proteins protect microtubules during simulated ischemia in cardiac myocytes. Am J Physiol 1998; 44: H2243–H2249. M JL, M R, HD R, B LL, D WH. Small heat shock proteins and protection against ischemic injury in cardiac myocytes. Circulation 1997; 96: 4343–4348. R PS, M JL, S EA, O H, D WH, D DK. Transgene overexpression of
B crystallin confers simultaneous protection against cardiomyocyte apoptosis and necrosis during myocardial ischemia and reperfusion. FASEB Journal 2001; 15: 393–402. E P, A WI, M JIA, H DJ, S MJ. Ischemic preconditioning: A potential role for constitutive low molecular weight stress protein translocation and phosphorylation? J Mol cell Cardiol 2000; 32(6): 961–971. I H, O K, N H, I T, K K. Phosphorylation of alpha B crystallin in response to various types of stress. J Biol Chem 1997; 272: 29934–29941. G N, N W, K R, H P, S W, D D. Ischemia-induced phosphorylation and translocation of stress protein alpha B crystallin to Z lines of myocardium. Am J Physiol 1998; 43: H1457–H1464. L SA, K AE. Protein phosphatase inhibitors and heat preconditioning prevent HSP27 dephosphorylation, f-actin disruption and deterioration of morphology in ATP-depleted endothelial cells. FEBS Letters 1998; 433: 294–300. G J, L H, G G, L JN, H J, L J. Regulation of actin filament dynamics by p38 MAP kinase-mediated phosphorylation of heat shock protein 27. J Cell Sci 1997; 110: 357–368. H J, H F, M F, L J. Oxidative stress-induced actin reorganization mediated by the p38 mitogen-activated protein kinase heat shock protein 27 pathway in vascular endothelial cells. Circ Res 1997; 80: 383–392. N MA, H ML, Q YZ, L KE, K

23.

24.

25.

26.

27.

28.

29.

30. 31.

32.

33.

34.

35.

RC. Delayed preconditioning of cultured adult rat cardiac myocytes: role of 70 and 90 kKa heat stress proteins. Am J Physiol 1997; 42: H861–H868. M MS, L DS, W JM, Y DM. Cardiac stress protein elevation 24 hours after brief ischemia or heat-stress is associated with resistance to myocardial-infarction. Circulation 1993; 88: 1264–1272. S JZ, T XL, K AA, P SW, Q YM, B R. Late preconditioning against myocardial stunning – an endogenous protective mechanism that confers resistance to postischemic dysfunction 24-h after brief ischemia in conscious pigs. J Clin Invest 1995; 95: 388–403. L H, C SJ, B AF, G A, L J. HSP27 multimerization mediated by phosphorylation-sensitive intermolecular interactions at the amino terminus. J Biol Chem 1999; 274: 9378– 9385. R T, E M, P X, K A, L G, D C, P C, W M, A A-P, B J, G M. Regulation of HSP27 oligomerization, chaperone function, and protective activity against oxidative stress/tumor necrosis factor alpha by phosphorylation. J Biol Chem 1999; 274(27): 18947–18956S. K K, I H, K K, I Y, I I, S S. Phosphorylation of alpha beta-crystallin in mitotic cells and identification enzymatic activities responsible for phosphorylation. J Biol Chem 1998; 273: 28346–28354. G N, H P, N W, K R, S W, D D. Binding of the stress protein alpha B crystallin to cardiac myofibrils correlates with the degree of myocardial damage during ischemia/reperfusion in vivo. J Mol Cell Cardiol 1999; 31: 569–580. A SC, S CL, G CE. Differential translocation or phosphorylation of Alpha B Crystallin cannot be detected in ischemically preconditioned rabbit cardiomyocytes. J Mol Cell Cardiol 2000; 32(7): 1301–1314. B IJ, MM DR. Stress (heat shock) proteins - molecular chaperones in cardiovascular biology and disease. Circ Res 1998; 83: 117–132. A H, A Y. Chaperone activity of alpha B crystallin suppresses tubulin aggregation through complex formation. Cell Struct Func 1997; 22: 539– 544. V LL, L E, B J, C ZI. Modulation of the swelling-activated amino acid channel of leishmania major promastigotes by protein kinases. Mol Biochem Parasitol 1997; 90: 449–461. B CP, L GS, B M, C SD, C MV, D JM. Ischemic preconditioning depends on interaction between mitochondrial KATP channels and actin cytoskeleton. Am J Physiol 1999; 45: H1361–H1368. S C, H ML, J RB. Cytoskeletal damage during myocardial-ischemia – changes in vinculin immunofluorescence staining during total in vitro ischemia in canine heart. Circ Res 1987; 60: 478–486. S C, J RB. Effect of cell swelling on plasma membrane integrity during anoxia in dog heart. Fed Proc 1984; 43: 294.


B Crystallin Translocation and Phosphorylation 36. G C, A S. Ischemia and the myocyte cytoskeleton – review and speculation. Cardiovasc Res 1993; 27: 1387–1403. 37. S D, E K, C DG, C P, G M. Identification of MAPKAP kinase-2 as a major enzyme responsible for the phosphorylation of the small mammalian heat-shock proteins. FEBS Letters 1992; 313: 307–313. 38. H HE, T DJ, M JJ, G CC. Alpha B Crystallin Gene Induction and Phosphorylation by MKK6-activated p38: A Potential Role for alpha B Crystallin as a Target of the p38 Branch of the Cardiac Stress Response. J Biol Chem 2000; 275(31): 23825–23833 (in press). 39. T A, B NS, E DT, E RM, D DK. Alpha-1-adrenergic receptor agonist-induced preconditioning in isolated working rat hearts. J Pharmacol Expl Therap 1995; 273s(2): 689–694. 40. C A, S PH. The p38-MAPK inhibitor, SB203580, inhibits cardiac stress-activated protein kinases/C-JUN N-terminal kinases (SAPKS/JNKS). FEBS Letters 1998; 426: 93–96. 41. B G, H DJ. Role of protein kinase C in ischemic preconditioning: player or spectator? Circ Res 1996; 79: 627–630.

1671

42. C M, L S, L U. Cardiac alphacrystallin. 3. involvement during heart ischemia. Mol Cell Biochem 1990; 97: 129–136. 43. M P, H E, W LA, A A-P. Large unphosphorylated aggregates as the active form of HSP27 which controls intracellular reactive oxygen species and glutathione levels and generates a protection against TNF alpha in NIH-2T3-ras cells. Biochem Biophys Res Comms 2000; 241: 187–192. 44. B WC, C Y,  R M,  R L,  J WW. Negative charges in the C-terminal domain stabilize the alpha B crystallin complex. J Biol Chem 1998; 273(43): 28085–28090. 45. B IJ, K J, X X, B D. Dynamic changes in O-GlcNAcylation and phosphorylation of cardiac
B-crystallin after heat stress and ischemia. Circulation 1999; 100(18): 57. (Abstract). 46. R EP, C MR, C RJ, H GW. Dynamic O-GlcNAcylation of the small heat shock protein alpha B crystallin. Biochemistry 1996; 35: 3578–3586. 47. M JL, M R, H HP, D WH. Small heat shock proteins and protection against ischemic damage in cardiomyocytes. Circulation 1998; 98(17): 620. (Abstract).