Simultaneous Determination of Tryptophan and Tyrosine by Spectrofluorimetry Using Multivariate Calibration Method

Printed in the Republic of Korea

   Tryptophan Tyrosine  

*

  

(2002. 2. 14 )

Simultaneous Determination of Tryptophan and Tyrosine by Spectrofluorimetry Using Multivariate Calibration Method

Sang Hak Lee*, Ju Eun Park, and Bum Mok Son

Department of Chemistry, Kyungpook National University, Taegu, 702-701, Korea

Received February 14, 2002)

.      (principal component regression, PCR)  (partial least squares, PLS)  !"#(tryptophan and tyrosine) $% &'( )  *+,.

!"# -./  01234 5678 257 nm9 :& ;&+,. < => !"# ?9 ,@ AB9 -.CD E( 32F %GH 280 nm~500 nm IJ? 0K235 LM:, N  PCR PLS 

OP LM,. < => !"# ?9 ,@ AB9 QRS 6 F TU& %G5 01235 V? 

OP W. U&X J TU& %G ABN Y#+,. Y#S ABN  relative standard error of prediction(RSEPa)N LM:, Z4 )[9 overall relative standard error of prediction(RSEP^m)B *+,.

: ,\' , Tryptophan, Tyrosine

ABSTRACT. A spectrofluorimetric method for the simultaneous determination of amino acids (tryptophan and tyrosine) based on the application of multivariate calibration method such as principal component regression and partial least squares (PLS) to luminescence measurements has been studied. Emission spectra of synthetic mixtures of two amino acids were obtained at excitation wavelength of 257 nm. The calibration model in PCR and PLS was obtained from the spectral data in the range of 280-500 nm for each standard of a calibration set of 32 standards, each containing different amounts of two amino acids. The relative standard error of prediction (RSEPa) was obtained to assess the model goodness in quantifying each analyte in a validation set. The overall relative standard error of prediction (RSEPm) for the mixture obtained from the results of a validation set, formed by 6 independent mixtures was also used to validate the present method.

Keywords: Multivariate Calibration, Tryptophan, Tyrosine

Tryptophan(Trp) tyrosine(Tyr) ]4 !"#4 ^ _ `? abc de c,.¹ Trp( fg h i$/ ED? j !"#k, lm na o>

 abc p :, iqrs, tu, v wxs

B oyc p c,. ib ( z{|

Try }| C~ €r XB c,. Tyr

Trp ‚c ƒ„( &…Wf †‡ˆ ., ‰  8 lŠ ‹(,.

Œc < => !"# &'( ,_c 

W Ž5 9 5~ X W (electro- chemical analysis),^2-5‘’  (atomic absorption spectrophotometry, AAS),⁶ :“ Hn”9•–—˜™

(high performance liquid chromatography, HPLC),^7-9

š    ] 5  E,. 5 a?  

4 ›B= œ ? 4 AB IJž> WŸ  E :,  ¡›¢ IJ= £,.^10-13 —Œ¤ ¥ˆ a 

… /ˆ =ž¦ 78?   §( /ˆ ¨

©| C~ 0123 aª[9 f … /

ˆ«   ;&X D¬,.^14-16 Œc †­ ®

¯X Jc )[9 Ž( °±²³ ´YW ,\

'  µ@ ¶B9 l: E,.^17-19 X¨ )

4  )·78 )·¸XN ;& U&¹¡

 LD? ABN Y#( º~ ,\' 4 ,_c 78? L4 0123  …

/ˆ ABN Y#X Jc C»¤, ;&S 0 123 y¼W[9 VX Jc )[9 S ,. ½¾?, ,\' 4 Œ => )VN /ˆ5

¤ … /ˆ5 c )· ¿À aª

ÁD¤B ÂÀ¤ À& à … /ˆ5

¡ÄW[9 e  E,. Œc ,\' 4

T¡-=%¡  (uv-visible spectrophotometry), WT¡  B(infarared spectroscopy), ¾« 

(raman spectroscopy), 0ÅÆÇ  (electron spin resonance spectrometry), ÈXÆÇ  (nucler magnetic resonance spectroscopy) ]^20-21 ¢:

E,.

É *?( Trp³ Tyr $%&'X J 

 [9 01235 L4 Ê  

(principal component regression. PCR)  

(partial least squares, PLS)  0123 ˲N ÂÀ( U&¹¡ L( )  

*+,. !"# -./  01234 5678

 257 nm9 :&:, )·78 280 nm~500 nm IJ

? LM,. L4 OP W. U&X J

 < => !"# -.H ?9 ,@ AB9 Q RS 6F U& %G5 01235  A BN Y#+:, Y#S ABN  relative standard error of prediction(RSEPa)N Y#+:, Z 4 )[9 overall relative standard error of prediction (RSEPm)B Y#+,.



. Trp³ Tyr4 Sigma(St. Louis, MO, USA)z 9² *Ì+,. ÍÎ zc ÏÏ Trp³ Tyr Ð8H4 Á&' Trp³ TyrN ÑÒ Ó A B= 1.0Ô10⁻³M CBÕ Öt+[k, pH 9.5 ׂ

H[9 zS boric acid-borateH4 H3BO3 6.18 g NaOH 2.8 g rØ 1 L Ó? Öt+,.²² rØ ( Milliporez(Bedford, MA, USA) Mill1-Q water system V L4 18.2 MΩ … ÐÙ =s Ñ

ÒN z+[k, OÚ H4 ÍÎ   Öt

 z+,.

. É ÍÎ? zc   Y( SPEXz (Edison, NJ, USA)? ÖpS Model FL111 Spectro- fluorimeter z+,. ‘[9( 450 W Xe Lamp, XÛU·X(reference detector)9( silicon diode U·

X, %G   ;&( l U·X(emission detector) 9( Hamamatsuz(Bridgewater, NJ, USA) Model R928 PMTN z+,. ;&78 1 nmÜ r=%Ý~?

1 h $Þ W ˲N LM:, &S 0123

 LX JV acquisition modeN s/r9 ¡Ä+,. % G pHN ;&X JV? Mettlerz(Schwerzenbach, Switzerland) Model MA235 pH/Ion analyzerN z

+,.

 . Trp, Tyr -.H )· 0123

;&X J zc H4 ,ß Z «5M,.

ÏÏ 1.0Ô10⁻³M~1.0Ô10⁻⁶ M Trp³ Tyr H Ö tc Ê 10 mL 'à¾0” Ötc ÏÏ Trp

³ Tyr H Á&' á: -.c Ê pH 9.5 boric acid - borate ׂH[9 ⡞> ãä? âÛH

 TU& %GH Öt+,.  H 1 cm Ô1 cm ^ å æ:, 56 78 257 nm9 

)· 78 çèD =~?  0123 LM,. 

Œc )[9 L4  0123 GRAMS32(Galactic Industries Co., NH, USA)? éD5 ÂÀ+,.

   . (principal component regression, PCR)^20-214 0123 G5

  ê  ­N *:, A B³ *c  ­ zN  OP

 *( ),. 0123 êë … /ˆ

AB êë  ÍêX J 0 123 êë  (i (1)) ê+:,

Y#S  ­(principal component score)5 A Bêë z )&i *( &(i (2)) Í ê+,.

(1) (2)

X( 0123 êë, T(  ­ êë, P^t 

êë, EX( 0123ìí(residual spectra) êë,. Y(

ABêë, B( Y êë(regression coefficient)k EY(nÔm)( ABìí êë,.

 (partial least squares regression,

PLS)^20-214 AB³ 0123 G5  $%

 ê AB  ­³ 0123  G  ­ zN  OP *

( ),. 0123 êë … /ˆ AB êë   OP *X J 0 123 êë AB êë $% (i (3), i (4))+:, *c <  ­5 z §îY (inner relation, B)N * OP *( &(i (5)) Íê+,.

(3)

(4)

(5)

X( 0123 êë, T( 0123 êë  ­  êë, P^t  êë, E^X( 0123ìí êë, Y(

AB êë, U( AB êë  ­ êë, Q^t( A B êë  ­ êë, EY( ABìí êë:, E( OP ìí,.

W OP *X J íU&(cross validation)& ê OP *( Ë 

e W  N ï&+,. ;S AB³ ÍÖ ABN    ½@ prediction residual error sum of squares(PRESS, i (6))N *

+,.

(6)

n4 ð9 *S %G , i( … /ˆ

, C^pred( OP  *c ?… /ˆ

ABk C^act4 ÍÖ AB,. *c PRESS
F-U&

 ´ PRESS \ = Ã,: e  E( 

 N OP *X Jc W  9

+,.

…­(outlier) U·X J 0123ìíN 

 Mahalanobis »À(d)N Y# ñò ê

»¤, F-ratioN ç9 »¤ F-ratioN 

F-U& Íê …­ U·+,.

(7)

(8)

d( Mahalanobis »À, R4 ìíêë, R( óôìí êë, Σ( #-Æ# êë,. n4 %G , ri³ r^j ( Ï %G ABìí ,. Mahalanobis »ÀN 

e õ( Mahalanobis »À óô aö[9 âÛ÷

í 3ø Vù( »À, ú »À E( %GN

…­[9 +:, F-ratioN ç9 (
( F- ratio= 1, ú ­ …­[9 +:, F-ratioN 

 F-U& êe õ( oÛ αN 0.019 

…­ U·+,.

OP W. U&X J TU&

(externa validation) %G a … /ˆ c relative standard error of prediction(RSEPa)N *+

:, … /ˆ n relative standard error of prediction(RSEPm) *+,. RSEP^a³ RSEP^mN *

( i ,ß ¤û§M,.

(9)

(10)

k( TU& %G , n4 … /ˆ ,

Cfound( OP  *c TU&¸2 A

Bk C^add4 TU& %G ÍÖAB,.

X=TP^t+E_X Y=TB E+ _Y

X=TP^t+E_X

Y=UQ^t+E_Y

Y=TBQ^t+E

PRESS (c_act–c_pred)²

i 1=

∑

=

d⁼(R R– )^t

∑

F ratio (n 1– )ri

r_j

i j≠

---

∑

= –

RSEP_a 100

C_found–C_add

( )

k 2

q 1

∑

C_add

k 2

q 1

∑

---

×

=

RSEP_m 100

C_found–C_add

( )

n 2

r 1= k

∑

q 1

∑

C_add

n 2

r 1= k

∑

q

∑

---

×

=

  

 . Trp Tyr )· 0123 ;&

X J ÏÏ AB= ,@ Trp Tyr -.H 

 56 78 257 nm :&%Ý: L4 )· 0 123 Fig. 1 ¤û§M,.

Fig. 1? ü  Eý 300 nm?( Tyr )·

¿À= ¤ûþ:, 350 nm?( Trp )· ¿À

= ¤ûþ,. ÿc < !"#  01235 a ª ÁD¤X õ
 X¨ &')[9 < !"

# $%&' D¬,(  ,.

PCR . OP ze  ­5 *

X J   +:,  ½@ : o(eigenvalue) Fig. 2 ¤û§M,.

Fig. 2? ü  Eý   (PC1) = 8 ú :o =>:, ,ß[9 <  (PC2)

 ú :o =s,.   = r=eÕ :o ­­ D(    E,. ( :o

 p4 54   zc G5 

ÇR ED? XB= p,(  !c,.

W PCR OP *X J íU& 

& ê PRESSN * :, — ïN Fig. 3

¤û§M,. Fig. 3? ü  Eý ¤ 

  õ PRESS= 2F    õ

 PRESS, p:, 3F    õ

PRESS( ,% r=+,. 4F    õ PRESS( ”| ›+:, 5F  

 õ PRESS( r=N ¤û§M,. 6F 

   õ PRESS( ,% ›+:, 7F?

² 10F   *c PRESS( ­íW f ›N ¤û§M,. —Œ¤ 10F … 

  õ *c PRESS( ­­ r=+,. Œc PRESS5  F-U&(α=0.01) ê+ õ PRESS \ = Ã,: e  E(  ( 6

M,.

OP? …­(outlier) U·X J 0 123ìíN *+:, — ïN Fig. 4 ¤¤§M ,. *c 0123ìíN  Mahalanobis»ÀN Y# ñò ê+:, F-ratioN Y#+:, F-U& ê+,. Fig. 4? ü  Eý 19 % G= =8 ú 0123ìíN ¤û§M:  %G(

Trp³ Tyr AB= ÏÏ 3.30Ô10⁻⁷ M9 4 %G+

,. Fig. 4 ¤û 0123ìíN  Mahalanobis Fig. 1. Fluorescence spectra of thirty eight mixtures contain-

ing different amounts of tryptophan, tyrosine in aqueous media: excitation wavelength, 257 nm; pH, 9.5.

Fig. 2. Plots of eigenvalue versus the PC number.

Fig. 3. Plots of PRESS versus the numbers of PC from cross validation analysis of the 32 sample set for PCR model.

»ÀN Y# ñò ê+ õ …­[9 U

·S %G( 19 %GM:, F-ratio= 1.0 … 

 =>(  …­[9  õ U·S …­4 19, 26 %GM:, Y#S F-ratioN  F- U& ê+ õ …­4 U·C> ,. ½¾

?, …­ U·) a? =8 ›c 4 F-ratio N c ) ,.

PLS1 . Trp &' Jc W PLS1 

OP *X J íU&(cross validation)&

ê PRESSN * :, Fig. 5 ¤û§M,.

Fig. 5? ü  Eý ¤    õ PRESS= 2F    õ PRESS

, ”:, 3F š 4F    õ

PRESS( ,% r=+>« 5F  ~

,% ú ›N +,. PRESS(  = r=

eÕ r=³ ›N »~?  = 17Á õž> ›,=  = 17?² 20ž>

?? r=c,.  = 20Á õ²( Á&c

 =s,. Fig. 5 ¤û PRESS5  F- U&(α=0.01) ê+ õ PRESS \ = Ã,

¾: e  E(  ( 5M,.

Tyr &' Jc W PLS1 OP *X J íU&& ê PRESSN * :, — ïN Fig. 6 ¤û§M,. Fig. 6? ü  Eý

 = 5Á õž> | ›,= 6²

| r=+,.  = 16Á õ²( Á

&c  ¤û,. Fig. 6 ¤û PRESS5 

 F-U& ê+ õ PRESS \ = Ã,¾ : e  E(  ( 4M,.

Trp³ Tyr c PLS1 OP? …­ U

·X J 0123ìíN * :, — ïN Fig.

7 Fig. 8 ¤û§M,. Fig. 7? ü  Eý 19

 %G= =8 ú 0123ìíN ¤û§M,. 19

%G( Trp AB= ÏÏ 3.30Ô10⁻⁷M9 4 %G+

,. Fig. 7 ¤û 0123ìíN  Mahalanobis

»ÀN Y# ñò ê+ õ …­[9 U·S %G( 19, 26 %GM:, F-ratio= 1.0

…  =>(  …­[9  õ U·S

…­4 19 %GM:, Y#S F-ratioN 

F-U& ê+ õ U·S …­4 ÃM,. Fig.

8?( 19 %G= =8 ú 0123ìíN ¤û§

M,. 21 %G( Trp³ Tyr AB= ÏÏ 3.30Ô10⁻⁷ Fig. 4. Plots of spectral residual versus sample number.

Fig. 5. Plots of the calculated PRESS using the PLS1 model of Trp versus the numbers of PC from cross-validation anal- ysis of the 32 samples.

Fig. 6. Plots of the calculated PRESS using the PLS1 model of Tyr versus the numbers of PC from cross-validation anal- ysis of the 32 samples.

M9 4 %G+,. Fig. 8 ¤û 0123ìíN

 Mahalanobis»ÀN Y# ñò ê

+ õ …­[9 U·S %G( 19, 26 %G

M:, F-ratio= 1.0 …  =>(  …­ [9  õ U·S …­4 19 %GM:, Y#

S F-ratioN  F-U& ê+ õ U·S

…­4 ÃM,.

PLS2 . PLS2 & W OP *

X J íU& ê PRESSN * :, Fig.

9 ¤û§M,.

Fig. 9? ü  Eý  = r=eÕ PRESS( ­íW[9 ›R   E:, 

= 8Á õ² PRESS= Á&c  =s,. Fig. 9

 ¤û PRESS5  F-U& ê+ õ

PRESS \ = Ã,¾: e  E(  ( 7

M,.

PLS2 OP? …­ U·X J 0

123ìíN * :, — ïN Fig. 10 ¤û§M ,. Fig. 10? ü  Eý 17 %G= =8 ú 0 123ìíN ¤û§M,. Fig. 10 ¤û ïN 

 Mahalanobis »ÀN  ñò ê

+ õ …­[9 U·S %G( 17 %GM:, F-ratio= 1.0 …  =>(  …­[9

 õ U·S …­4 17 %GM:, Y#S F- valueN  F-U& ê+ õ U·S …

­4 ÃM,.

. OP U& J T U&¸

2 ÍÖAB³ OP  … /ˆ

 ABN Y#+:, &'ï &BN ¤û§(

Fig. 7. Plots of the calculated spectral residual using the PLS1 model of Trp versus sample number.

Fig. 8. Plots of the calculated spectral residual using the PLS1 model of Tyr versus sample number.

Fig. 9. Plots of the calculated PRESS using PLS2 model ver- sus the numbers of PC from cross-validation analysis of the 32 sample set for PLS2.

Fig. 10. Plots of spectral residual versus sample number to PLS2 model.

RSEPa³ RSEPmB Y#+[k — ïN Table 1

Table 2 ¤û§M,.

TU&¸2( ;&c 38F 01235 a?  pJ9 6FN · *+,. Table 1( Trp

&'ïN ¤û§M,. Table 1? ü  Eý

PCR  *c … /ˆ RSEPa 8.2%9

=8 ú  =>:, PLS2  *c … /

ˆ RSEP^a 5.6%9 í= =8 pß   E,.

Table 2( Tyr &'ïN ¤û§M,. Tyr


B Trp³ Z4 ïN Û,. PCR  *c 

… /ˆ RSEP^a= 8.9%9 =8 ú  =>:

PLS2  *c … /ˆ RSEPa= 3.8%9

í= =8 Wß   E,. n… /ˆ

&'ï c OP &BN ó=X J

RSEPm *+,. PCR  *c n 

… /ˆ RSEP^m4 8.6% M:, PLS1  * c n … /ˆ RSEP^m4 5.1% M[k,

PLS2N  *c n … /ˆ RSEPm4

4.6%M,. RSEP^m ïB •=> PLS2= =8

í= Wß ¤û§: E,.

 

Trp³ Tyr  0123 ?9 >« ,\'

f PCR, PLS1š PLS2N  Trp³ Tyr

&' e  E,. 0123ìíN  ÏÏ )

? Mahalanobis »À c ñò, F-ratio

š F-U&[9 …­ U·+ õ F-ratioN 

( ) =8 ›c )M,. Ï U&)[9 U&%G ABN ;:, RSEP^a³ RSEP^mN *

 õ PLS2  ;c Trp³ Tyr AB= =8 p4 RSEP^a³ RSEP^m ¤û§M,.

   

1. Davies, J. S. Amino Acids and Peptides; Chapman &

Hall, London, 1985.

2. Vahjen, W.; Bradley, J.; Biliteuski, U.; Schmid, R. D.

Anal. Lett. 1991, 24, 1445.

3. Villarta, R. L.; Cuningham, D. D.; Guilhault, G. G. Tal- anta 1991, 38, 49.

4. Tsai, H.; Weber, S. G. Anal. Chem. 1992, 64, 2897.

5. Saurina, J.; Hern ndez-Cassou, S.; F bregas, E.; Alegret, S.

Anal. Chim. Acta. 2000, 405, 153.

6. Song, J.; Wang, X.; Xu, F.; Jiangyan, L. Huaxue Fence 1991, 27, 225.

7. Watanabe, A. J. Chromatogr. Biomed. Appl. 1992, 121, 241.

8. Polanuer, B.; Sholin, A.; Demina, N.; Rumiantsova, N.

J. Chromatogr. 1992, 594, 173.

9. Wills, R. B. H.; Wimalsairi, P.; Greenfield, H. J. Micro- nutr. Anal. 1985, 1, 23.

10. Yang, J.; Zhao, W. Anal. Lett. 1993, 26, 2291.

11. Jie, N.; Yang, J.; Zhan, Z. Anal. Lett. 1993, 26, 2283.

12. Damiani, P.; Ibanez, G. Olivieri, A. Anal. Lett. 1993, 26, 247.

13. Mohamed, F. A. Anal. Lett. 1995, 28, 2491.

14. Konstantianos, D. G.; Ioannou, P. C.; Efstathiou, C. E.

Analyst 1991, 116, 373.

15. Konstantianos, D. G.; Ioannou, P. C. Analyst 1992, 117, 877.

16. Gorges, J.; Ghazarian, S. Anal. Chim. Acta. 1993, 276, 401.

17. Soper, S. A.; Warner, I. M.; McGown, L. B. Anal.

Chem. 1998, 70, 477.

18. Giamarchi, P.; Stephan, L.; Salomon, S.; Le Bihan, A.

J. Fluorescence 2000, 10, 393.

19. Guiteras, J.; Beltr n, J. L.; Ferrer, R. Anal. Chim. Acta.

1998, 361, 233.

Table 1. Analytical results for tryptophan obtained by applying PCR, PLS1 and PLS2 to fluorescence spectral data^a

Actual Predicted

PCR PLS1 PLS2

3.30×10⁻⁵ 3.30×10⁻⁴ 3.30×10⁻⁵ 3.30×10⁻⁶ 3.30×10⁻⁴ 3.30×10⁻⁴

3.17×10⁻⁵ 3.64×10⁻⁴ 3.92×10⁻⁵ 3.93×10⁻⁶ 3.12×10⁻⁴ 3.04×10⁻⁴

3.35×10⁻⁵ 3.41×10⁻⁴ 3.45×10⁻⁵ 3.24×10⁻⁶ 3.36×10⁻⁴ 3.39×10⁻⁴

3.34×10⁻⁵ 3.46×10⁻⁴ 3.48×10⁻⁵ 3.46×10⁻⁶ 3.05×10⁻⁴ 3.20×10⁻⁴

aRSEPa: PCR, 8.2%; PLS1, 5.7%; PLS2, 5.6%.

Table 2. Analytical results for tyrosine obtained by applying PCR, PLS1 and PLS2 to fluorescence spectral data^a

Actual Predicted

PCR PLS1 PLS2

3.30×10⁻⁴ 3.30×10⁻⁴ 3.30×10⁻⁶ 3.30×10⁻⁴ 3.30×10⁻⁵ 3.30×10⁻⁴

3.58×10⁻⁴ 2.87×10⁻⁴ 2.19×10⁻⁶ 3.04×10⁻⁴ 2.10×10⁻⁵ 3.29×10⁻⁴

3.57×10⁻⁴ 3.27×10⁻⁴ 3.17×10⁻⁶ 3.32×10⁻⁴ 3.63×10⁻⁵ 3.42×10⁻⁴

3.39×10⁻⁴ 3.51×10⁻⁴ 3.90×10⁻⁶ 3.22×10⁻⁴ 3.67×10⁻⁵ 3.35×10⁻⁴

aRSEPa: PCR, 8.9%; PLS1, 4.5%; PLS2, 3.8%.

20. Martens, H. Naes, T. Multivariate Calibration; Wiley, New York, 1989.

21. Sharaf, M. A.; Illman, D. L.; Kowalski, B. R. Chemo-

metrics; Wiley, New York, 1986.

22. Hern ndez, M. J. M.; Cama as, R. M. V.; Cuenca, E.

M.; Alvarez-Coque, M. C. G. Analyst 1990, 115, 1125.