II‘HH

IIHI

l
i

ll

"“ H
03-h
0701

 

THS-

REMAR WE“ RWRRED ER‘JEJERAHBN
(3'? WE C RBGML FREQRERCY 0F RAD

‘i‘hesis for Rte Regree {If M. S.
at" Tiff SSW-SE E SERVERS“?
RR MICHEL RRRRECKE R
19.73

ABSTRACT

RAMAN AND INFRARED INVESTIGATION OF
THE CARBONYL FREQUENCY OF NAD

BY
Dan Michael Patrick II

The coenzyme NAD was studied by Raman and infrared
spectroscopy. This spectroscopic investigation was confined
to the carbonyl group of the nicotinamide moiety, which was
monitored to seek evidence regarding the possibility of an
intramolecular hydrogen bond existing between the amino
group of the adenine moiety and the carbonyl group of the
nicotinamide moiety of NAD.

The following compounds were chosen as models to
provide a basis for interpreting the spectrum of NAD: nico-
tinamide, 3-acetyl pyridine, nicotinaldehyde, the methylated
derivatives of these compounds, and nipecotamide.

It was concluded that, in hydrogen bonding solvents,

NAD+ exists with a significant ?- + resonance contribu-
C = NH
/ 2

tion. In nonhydrogen bonding solvents, this contribution is
decreased, and the carbonyl frequency has a more classical

value.

Dan M. Patrick

NAD+, NADH, and the aldehyde analog of NAD+ were
examined at room temperature and at 70°C. It has been re-
ported that at increased temperatures the intramolecular
stacking is destroyed and NAD is found in an open configur—
ation. The lack of a temperature dependence of the car-
bonyl frequency strongly suggests the absence of an intra-
molecular hydrogen bond between the amino group of the ade—
nine moiety and the carbonyl of the nicotinamide group of

NAD .

RAMAN AND INFRARED INVESTIGATION OF

THE CARBONYL FREQUENCY OF NAD

BY

Dan Michael Patrick II

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Biochemistry

1973

ACKNOWLEDGMENTS

A word of thanks goes to Dr. Farnum for the loan
of the IRTRAN cells and also to Mark Greenberg for the pur
rification of the dimethylsulfoxide used throughout this
study.

The interest and helpful discussions of many of the
faculty at Michigan State University is greatly appreciated.

The Molecular Spectroscopy group of the Chemistry
Department has provided a very friendly, helpful and re—
laxed atmosphere for this study.

A special thanks goes to Dr. George Leroi and Dr.
John Wilson for their interest, invaluable guidance, pa-
tience, encouragement and friendship with which they have
viewed this research. A special consideration goes to Dr.
Leroi, who is responsible for sparking this author's inter-
est in spectroscopic analysis and has been an inspiration
to his educational growth.

Finally, this author wishes to thank his parents

for their unfaltering support and encouragement.

ii

TABLE OF CONTENTS

INTRODUCTION I O O O O C O O O O O O O O O O O C I 0

Chapter

I. PRELIMINARY REVIEW OF THE CARBONYL FREQUENCY -

OF THE PYRIDINE MOIETY OF NAD+ . . . . . . .

Introduction . . . . . . . . . . . . . . .
Experimental . . . . . . . . . . . . . . .
Results and Conclusion . . . . . . . . . .

II. ANALYSIS OF THE CARBONYL FREQUENCY OF
QUATERNARY PYRIDINE MODEL COMPOUNDS . . . .

IntrOduc tion 0 O O O O 0 O O O O O O O O 0
Experimental . . . . . . . . . . . . . . .
Results and Conclusion . . . . . . . . . .

III. ANALYSIS OF THE CARBONYL FREQUENCY OF NAD .
Introduction . . . . . . . . . . . . . . .
Experimental 0 O O O O O O O O O O O I O 0

Results and Conclusion . . . . . . . . . .

LIST OF REFERENCES . . . . . . . . . . . . . . . . .

iii

26
26
27
34
36
36
38
46

50

Table

1.

LIST OF TABLES

Page

Positions of the assigned carbonyl frequencies
for nicotinamide, 3—acety1 pyridine, and
niCOtinaldehYde O O O O 0 O O O O O O O O O O O 20

Position of the carbonyl frequency for
nipeco tinamide O O O O O O O O I O O O O O O O 21

Effect of the solvent composition on the car—
bonyl frequency of nicotinamide . . . .A. . . . 22

Melting point values for N-methyl nicotinalde-
hyde iodide and N—methyl 3-acetyl pyridine
iOdide in 0C O O O O O O O O O 0 O O O O O O O O 30

Positions of the carbonyl frequency for N-

methyl nicotinamide iodide, N-methyl 3-acety1
pyridine iodide, and N-methyl nicotinaldehyde
iodide . . . . . . . . . . . . . . . . . . . . 33

Ultraviolet spectra of selected compounds . . . 39
Effect of temperature on the carbonyl frequency

in the Raman spectra of NAD and similar mole-
cules O O O O O O O O O O I O O O O O O O O O 42

iv

INTRODUCTION

Cofactors are nonprotein structures which many en;
zymes require for activity. The cofactor can be a metal
ion or a complex organic molecule, which is classified as
a coenzyme.

NAD is the symbol for the oxidoreduction coenzyme,
nicotinamide adenine dinucleotide, which is involved in the
transfer of hydrogen atoms and electrons. Figure l is the
NAD structure drawn with no consideration given to its con-
formation.

Interactions between the adenine and nicotinamide
moieties in NAD have been detected by various spectroscopic
techniques and evidence has been accumulated which indi-
cates that in aqueous solutions and at room temperature NAD
tends to exist in a folded or internally complexed form.1-7
The suggested structure seems to resemble the stacked struc-
tures of polynucleotides and nucleic acids.

Miles and Urryl examined NAD using circular dichro-l
ism and absorption spectroscopy. The analysis dealt with
the effects of interbase coupling on the electronic transi-

tions. The presence of reciprocal relations in the circu-

lar dichroism spectra gave strong evidence that the folded

 

RAJ

 

 

OHZC O
HOWE—O OH OH
O
|
HO—Pz-O
R
OH C O
2 .
When R is equal to OH OH
H
J
C: H
TIC: 4.,/’ F4
\\\\hJ///x:

NAD+

Figure 1. Structure of NAD.

 

*icfi <3
HC/ \ c

state is substantially populated at low temperatures. This
examination was performed on both alpha and beta NAD, which
are the two isomers of the coenzyme corresponding to the
alpha and beta conformation around the glycoside linkage.

As an experimental parameter is varied which brings
two different chromOphoric groups into juxtaposition, one
might expect that close lying transitions, in particular,
exhibit a coupling in which the circular dichroism peak
due to a transition in the first chromophore becomes more
positive while the circular dichroism peak of a close-lying
electronic transition in the second chromophore becomes
more negative in a reciprocal manner. This coupling has
been called "reciprocal relations" in optical rotation.8

Scott 33 al.2 examined the fluorescence properties
of NADH and model compounds in which the linkage of the
adenine and the dihydronicotinamide moieties of NADH was
constructed with tri- and hexamethylene chains. They con-
cluded that in water NADH behaves like a molecule in which
the adenine and dihydronicotinamide moieties are proximate,
as in the trimethylene model compound. In nonaqueous media
NADH resembled the hexamethylene model compound, in which
the terminal heterocyclic groups are remote.

Secrist and Leonard3 constructed an "abbreviated"
model system of NAD+ which contained the elements of the
two heterocyclic rings and the sugar. These compounds were

constructed to provide an improved model system which more

closely resembles NAD than those with variable length meth-
ylene groups between the adenine and the nicotinamide moi-
eties for investigating the base interactions. The "abbre-
viated" model systems and NAD were studied by ultraviolet
spectroscopy and circular dichroism. The ultraviolet hypOF
chroism of the model systems indicated the presence of an
intervring interaction which was greater than the one shown
for NAD+. The spectroscopic techniques employed verified
the existence of an intramolecular interaction in aqueous
media between the quaternized nicotinamide ring and the
adenine ring.

Nuclear magnetic resonance spectroscopy has been
employed to study NAD and the following observations have
been made: (1) There are differences in the chemical shifts
of the adenine and the nicotinamide protons as a function

of the pH.4

(2) The C-2 and the C-6 protons of the nico-
tinamide moiety have a differential shielding and it is
stated that they are symmetrically disposed with reference

5

to the N-glycoside bond. (3) NADH has an NMR spectrum

which contains an AB quartet.6

(4) The NMR spectrum of N—
methyl-N-ethylnicotinamide adenine dinucleotide has four
N-methyl resonances.7

From the above NMR data, Sarma and Kaplan have pro-
posed a helical model for NAD. The identification of the
specific conformation of the folded form by NMR spectroscopy

and the proposed Sarma—Kaplan model are still undergoing

critical analysis.9

Adams gt al.10 have analyzed the lactate dehydro-
genase—NAD complex using an X-ray structural determination.
The NAD associated with the enzyme was in an essentially
linear array.

NAD has complete torsional freedom over the entire
linkage from the adenine moiety to the pyridine moiety of
the molecule. Thus, a question to be considered is: Does
an intramolecular hydrogen bond exist between the amino
group of the adenine moiety and the carbonyl of the nico-
tinamide moiety which would be responsible for the folded
conformation and complement the base stacking?

An intramolecular hydrogen bond occurs only when a
proton donor and a proton acceptor site on the same mole—
cule are in a favorable spatial configuration; that is, the
distance between the hydrogen of the donor group and the
acceptor site is between 1.4 and 2.5 A, and the angular
orientation of the acceptor site does not deviate greatly
from the bond axis of the donor group, A-H.11

The nature of hydrogen bonding can be examined by
spectroscopic techniques. From them, one can obtain evi-
dence relating to the formation of a weak bond, the involve-
ment of a specific covalently bound hydrogen atom, and the
participation of a specific acceptor group. With this back—
ground information, it was decided to employ laser-Raman
techniques to monitor the carbonyl group in NAD as a possi-

ble acceptor site. The vibrational modes of the hydrogen

bond acceptor, in this case C=O, can be shifted by hydrogen
bonding. These shifts are always to lower frequencies.12

The analysis of the carbonyl stretching frequencies
in various types of carbonyl compounds suggests that the
observed band position results from the interplay of sev-
eral factors, among which the following are clearly impor-
tant: (1) Physical state of the compound, (2) Inductive
effects, (3) Electronic and mass effects of neighboring
carbonyl groups, (4) Hydrogen bonding, (5) Enolization, and
(6) Solvent effects.

Raman spectroscopy is based on the interaction of
light with a molecule resulting in inelastic photon scatter-
ing. Vibrational Raman scattering occurs if there is a con—
comitant change in the polarizability of the molecule, that
is when the interaction of the electromagnetic field with
the vibrational motion produces an induced dipole moment.
There is low interference of water compared to infrared
spectroscopy. Therefore, aqueous solutions of biological
systems can be examined precisely in the state in which
they are found, although higher concentrations are normally
necessary since the second order Raman effect is rather
weak. Thus, as a check on positions and assignments, in-
frared spectroscopy can also be employed. Infrared is also
employed because the carbonyl stretching absorption band in
the region of 1870-1540 cml has a relatively constant posi—

tion, high intensity, and relative freedom from interfering

bands, making this one of the easiest bands to recognize

in infrared spectra.13

This thesis deals with the Raman and infrared spec-
troscopic study of the carbonyl frequency of NAD, with the

specific purpose being to seek evidence for or against the

formation of an intramolecular hydrogen bond.

CHAPTER I

PRELIMINARY REVIEW OF THE CARBONYL
FREQUENCY OF THE PYRIDINE MOIETY
OF NAD+

Introduction

 

In a vibrational spectroscopic study one must de-
termine whether the selected methods and techniques are ap—
plicable to the problem of interest.

Initially, nicotinamide, nicotinaldehyde, 3-acetyl
pyridine and nipecotamide were studied at room temperature
and in various solvent systems as models for the pyridine
moiety of NAD+. Figure 2 depicts the structures of the
compounds that were explored in this preliminary study.
The data obtained showed that Raman spectroscopy repre-
sented a potentially valuable method for studying the en-
vironment of the carbonyl group of NAD+. They also estab-
lished a basis for interpreting the Raman spectra of the

intact NAD+ molecule.

Experimental

 

Nicotinamide was obtained from the Sigma Company
and 3-acetyl pyridine, nicotinaldehyde and nipecotamide

were obtained from the Aldrich Chemical Company and were

Nicotinamide

3-Acetyl pyridine

Figure 2.

Nicotinaldehyde

Nipecotamide

Structures of compounds examined in Chapter I.

10

used without further purification. The solvents employed
were: p—dioxane from the Matheson, Coleman and Bell Com-
pany, deuterium oxide from the Columbia Organic Chemicals
Company, methyl alcohol-d4 from the British Oxygen Company,
and acetonitrile and dimethylsulfoxide from the J. T. Baker
Company. The dimethylsulfoxide was further distilled over
4A molecular sieves to insure the absence of any water.

The remaining solvents were used without further purifica-
tion.

The instrument used was a laser-Raman spectrometer14
consisting of: either a Spectra-Physics Model 164 Ar+ laser
(providing a maximum output of 1.6 watts of 5145 A radia-
tion) or a Spectra—Physics Model 165 Kr+ laser (providing a
maximum output of 0.76 watts of 6471 A radiation); the
Spectra-Physics 265 exciter; Spex 1400 double monochrometer;
RCA C31034 photomultiplier tube; Victoreen VTE-l D.C. amp—
lifier; and a Hewlett-Packard Moseley 7100 B strip chart re-
corder.

Typical Raman instrument settings using the Ar+
laser were: time constant of l or 3 seconds; detector temps
erature less than -20°C, and voltage between 1000-1300
volts; scan speed 10-2.5 R/min.; slits 100-200-100 microns;

10 amps; 5145 A exciting line with power out-

sensitivity 10-
put 1.0-1.3 watts; chart speed of 0.5-0.2 in/min. When the
Kr+ laser was substituted the following changes were made:

slits 200-200-200 microns; sensitivity 10’11 amps; 6471 R

ll

exciting line with power output 0.5-0.76 watts; photomulti-
plier voltage of 1000-1100 volts.

The cells employed for the Raman samples were four
millimeter o.d. glass tubing, which was first sealed at one
end, bent into an L—shape design, sample added, and then
sealed at the other end. The laser beam was focused into
the glass tubing by a lens (focal-length 10 cm.). The
Raman scattering was collected at 90° from the laser's in-
cident direction and focused on the monochromator entrance
slit. See Figure 3 for a schematic representation of the
Raman sample cell and the illumination and scattering op-
tics.

A Perkin-Elmer 225 infrared spectrometer was also
used. The IR spectra were all double beam. Aqueous solu-
tions were run in 0.2 millimeter thick cells with IRTRAN
windows, whereas the nonaqueous solutions were run in
0.103 millimeter thick cells with NaCl windows. The fol-
lowing IR instrument settings were employed: slit program
4; pen traverse time 3 seconds; gain between 100-500.

All solutions examined by Raman analysis were 0.25M
concentration except the nicotinamide solutions in aceto-
nitrile and p-dioxane and the aqueous nipecotamide. They
were 0.15M, 0.20M, and 1.0M respectively. The solutions
inspected using IR analysis were saturated solutions, ex-
cept the D20 solution of nicotinamide and the DMSO solu-

tions of nicotinamide and nipecotamide which were 0.25M.

12

   

 

 

*
* Laser Beam
*
'k
'k
*
it
* fl
/
* /
* /
/
* /
* L
.4
Sample. 4 mm
solution
to
Monochromator (transverse-transverse)

Figure 3. Raman cell.

13

Figures 4 through 9 are examples of the Raman and IR spec-
tra that were obtained. Table 1 lists the values of the
assigned carbonyl frequencies for nicotinamide, 3—acetyl
pyridine, and nicotinaldehyde in the various solutions.
Table 2 lists the assigned carbonyl frequencies for nipec0v
tamide in D20 and DMSO.

Solutions of 0.25M nicotinamide in varying percen-

tages of D O and p—dioxane by volume were prepared and were

2
examined by Raman analysis, and Table 3 lists the values
for the assigned carbonyl frequency.

The error limit or reproducibility of all assign-

ments is $1.0 cml for both Raman and IR values.

Results and Conclusion

 

Nicotinamide, 3-acetyl pyridine and nicotinaldehyde
are similar in that each contains a pyridine ring with a
carbonyl in the 3 position. As a result, any resonance or
inductive effects of the ring system on the carbonyl fre-
quency should be identical for the three compounds.

The above mentioned three compounds differ in the
R group substituted on the carbonyl group. Basically, none
and CH

of the R groups, H, NH present any major steric

2' 3’
hindrance to solvent interactions at the carbonyl position.
Consequently, in considering the electronic distribution of
the carbonyl group, any changes brought about by hydrogen

bonding of a solvent system should also be similar for the

three compounds.

l4

. -1
Wavenumber in cm

— 1800
1720
1640
1560
1480

-- 1400

—
u—
_
—

 

 

 

 

 

 

 

1690

 

Figure 4. IR spectrum of nicotinamide in chloroform.

15

1800
1720
1640
1560
P 1480
1400

I l I
Wavenumber in cml

h-

 

 

 

 

 

 

 

C

1690

Figure 5. IR spectrum of nicotinamide in p—dioxane.

Figure 6.

l6

 

 

 

l I J L, l I
O O O O O O
O N <‘ \D co 0
(I) \0 Ln fl' '4‘
H H H H .4

U _l
Wavenumber in cm
1633
r’ l I l I H

 

IR spectrum of nicotinamide in D20.

17

1691

 

 

 

 

. -l
Wavenumber in cm

 

1731.78-a
1700.39-4
1668.89-
1637.28—
1605.56+
1573.74-

Figure 7. Raman spectrum of nicotinamide in p-dioxane.

18

 

 

 

 

 

. -1
Wavenumber in cm

 

1731.78-
1700.39—-
1668.89-
1637.28-1
1605.56-4
1573,74_.

Figure 8. Raman spectrum of nicotinamide in 75 percent
p-dioxane and 25 percent D20.

l9

 

 

1637

. -1
A _- - A Wavenumber in cm

 

1668.89—

1731.78F
1700.39—
1605.56‘1
1573.74fi

Figure 9. Raman spectrum of nicotinamide in D20.

20

 

 

 

     
   

 

 

 

 

 

 

 

 

 

 

Table 1. Positions of the assigned carbonyl frequencies for
nicotinamide, 3—acety1 pyridine, and nicotinaldehyde.
Carbonyl Position in cm Solvent Pr0perties
Solvent R Group 1
‘ Dipole Die ectric
NHZ H I CH3 moment15 constant 16'17
Raman IR Raman
D20 1637 1633 1705 1684 1.88 (D) 78.25
CD3OD 1658 1654 1711 1693 2.97* 32.63*
1685
CH3CN 1689 1690 1707 1692 3.39 38.8
DMSO 1686 1687 1702 1688 3.9 (B) 48.9
p—
dioxane 1691 1690 1707 1692 0 2.2
H20 —- -- 1703 1685 1.92 78.54
C6H6 —- 1690 -- -- 0 2.28
CHCl3 —- 1690 -- -- 1.55 4.81
Pure
Cmpd 16755 -- 1700n 1688n ----------
3 solid
n neat
* Values for methanol, CH3OH
Error limits on all assignments, both Raman
and IR are i 1.0 cm
(B) in a benzene solution
(D) in a dioxane solution C)
II

Basic Structure is =

21

Table 2. Position of the carbonyl fre-
quency for nipecotinamide.

 

-i

 

 

Carbonyl Position in cm
Solvent Raman IR Comments
D20 1623 1621 1.0 M
DMSO 1672 1674 0.25 M

 

 

 

Error limits for Position 1 1.0 cml

22

Table 3. Effect of the solvent composition on the
carbonyl frequency of nicotinamide.

 

m

 

Percent Percent

of of Position Width at Half
p-dioxane 920 in cm Height in cm1

100 0 1691 11

94 6 1665 27

88 12 1659 27

75 25 1658 25

50 50 1648 37

25 75 1639 32

0 100 1637 32

 

Error limits:
Width at Half Height: 1 10% for each
indicated value.

Position: i 1.0 cml

 

23

Upon inspection of the data in Table 1 for 3-acety1
pyridine and nicotinaldehyde, one observes that the carbonyl

1 and 11 cml

frequency varies only 9 cm for the respective
compounds over the entire range of solvents employed. Thus,
one can conclude that the solvent has only a minor effect
upon the location of the carbonyl position.

When comparing the results of nicotinamide with
those for nicotinaldehyde and 3-acety1 pyridine, one ob-
serves an anomalous effect in a D20 or CDBOD solution. All
of the remaining solvents are within 5 oil of each other
and the agreement between the IR and Raman data is excellent.
As Table 1 indicates, this anomalous effect is apparently
not related to the dielectric constant or dipole moment
properties of the solvents. D20 and CD3OD differ from the
other solvents due to their hydrogen bonding properties.

In considering the question of an intermolecular hy-
drogen bond in nicotinamide, it must be pointed out that in
all cases only one carbonyl frequency was observed. Often,
such systems result in two carbonyl frequencies being ob-
served. One is due to a free and the other due to a hydro-
gen bonded carbonyl group.

Intermolecular hydrogen bonding involves associa-
tion of two or more molecules. The extent of intermolecular
hydrogen bonding is temperature dependent. Also, the bands
that result from intermolecular hydrogen bonding generally
disappear at low concentrations, less than about 0.01M in

nonpolar solvents.18

24

Figure 4 is an IR spectrum of a saturated CH3C1 so-
lution of nicotinamide, which is less than 0.01M. Figures
5 and 7 are the IR and Raman spectra of a 0.20M nicotin—
amide solution in p—dioxane. There is no change in the
carbonyl position. Thus, there is no evidence of a hydro-
gen bonded carbonyl resulting from an intermolecular hydro-
gen bond. Therefore, intermolecular effects should pre-
sent no problem when interpreting the other spectral data
obtained in this study, typically using concentrations of
approximately 0.25M.

When one considers Table 3, one observes that the
carbonyl frequency is continually lowered as D20 is added
to a dioxane solution of nicotinamide. Dioxane is con-
sidered to be an inert solvent because it shows no spec-
tral evidence of an interaction with the carbonyl group.
Therefore, the lowering of the carbonyl frequency is re-

lated to the presence of D O, which is a hydrogen bonding

1

2
solvent. This study confirms the assignment of 1637 cm

in an aqueous solvent as the carbonyl frequency.
Nipecotamide was also examined. The only double
bond is the carbonyl group in the 3 position. As a result,
there is no possible interaction with the ring system.
When Table 2 is examined and compared with the data for
nicotinamide, one observes the same solvent effect. That
is, the presence of a hydrogen bonding solvent results in

a major lowering of the carbonyl frequency when an amino

25

group is adjacent to it. This solvent effect is not pres-
ent with nicotinaldehyde and 3-acetyl pyridine. So, the
lowering of the carbonyl frequency is not due to the sol-
vent itself. It appears to be directly related to the pres-
ence of an adjacent amino group.

In terms of the adjacent amino group, one can write

the following resonance structure:13

0
I I

C
R/ \NH2 R/ \KIHZ

(I) (II)

The resonance effect increases the carbonyl bond
length and reduces the frequency of absorption.13 As a re-
sult of increased interaction (via hydrogen bonding) with
resonance form II, hydrogen bonding solvents could increase
the contribution of such an effect; this could provide an
explanation for the anomalous lowering of the carbonyl fre-
quency in a hydrogen bonding solvent. In a nonhydrogen
bonding solvent, form II would be expected to make a lesser

contribution to the electronic structure of the carbonyl.

CHAPTER II

ANALYSIS OF THE CARBONYL FREQUENCY
OF QUATERNARY PYRIDINE
MODEL COMPOUNDS

Introduction

 

In the previous chapter it was noted that in an
aqueous medium nicotinamide appears to exhibit a resonance
effect between the carbonyl group and the adjacent amino
group. NAD+ exists with a plus charge on the quaternary
pyridine ring. What must be considered next is to what ex-
tent a positive charge on the pyridine ring alters the car-
bonyl frequency.

In these experiments, N-methyl nicotinamide iodide,
N-methyl 3-acetyl pyridine iodide, and N-methyl nicotin-
aldehyde iodide were studied at room temperature and in
different solvent systems for the above purpose. NAD+ has
an N-glycoside bond to the quaternary pyridine ring. If
quaternization of the nitrogen in the pyridine ring does
result in delocalization of the electrons from the carbonyl
group, it should make no drastic difference as to the group
that provides the quaternary linkage to the pyridine ring.
Therefore, the methylated compounds of nicotinamide, nico-

tinaldehyde, and 3-acety1 pyridine were selected as model

26

27

compounds. Figure 10 is the structure of the compounds in-

spected in this phase of the study.

Experimental

 

N-methyl nicotinamide iodide was obtained from the
Sigma Company, methyl iodide was obtained from the J. T.
Baker Company, nicotinaldehyde and 3~acetyl pyridine were
obtained from the Aldrich Chemical Company and all were
used without further purification. The solvents employed
were deuterium oxide from the Columbia Organic Chemical
Company and dimethylsulfoxide from the J. T. Baker Company.
The dimethylsulfoxide was further distilled over 4A molecu-
lar sieves to insure the absence of any water.

The Raman spectrometer and cells used were described
in the previous chapter (see Chapter I, Experimental).

Nicotinaldehyde and 3-acety1 pyridine were methyl-

ated by the following reaction:19

O
\ (Ijl\ A \ C\R

 

/ 3 60C +/ 1...
N

(Liquid) (Liquid) 3 (Solid)

where R = CH3 and H

28

| NH2 |
+/ I’ +/ I-
i 1
CH3) CH3
N—methyl nicotinamide N—methyl nicotinaldehyde
iodide iodide
O
\ g
\\
CHa
+/ _
N I
|
CH3

N-methyl 3-acety1 pyridine iodide

Figure 10. Structures of compounds examined in Chapter II.

29

The synthesis utilized a four-fold excess of methyl
iodide. The reaction began almost immediately, with a yel-
low solid being formed in both cases. The reaction was
kept at 60°C. to allow the excess CH3I to evaporate. The
products were dried and recrystallized twice from spectral-
grade iSOpropyl alcohol (Matheson, Coleman, and Bell Com—
pany). The melting points were taken and are compared in
Table 4 with the literature values. The agreement is ex-
cellent.

The Raman spectra of the solid methylated model
compounds were taken using the Kr+ laser and the 6471 A ex-
citing line. The crystals were placed in melting point
capillary tubes and the open end was at 90° to the incident
laser light. Spectra were obtained for 0.25M solutions of

the three methylated compounds in D O and dimethylsulfoxide.

2
The other solvents used in the earlier study (Table 1) dis-
solved inadequate amounts to permit obtaining of Raman spec-
tra. The solutions were examined using the Kr+ laser and
the 6471 A exciting line, and in addition, the D20 solutions
were examined using the Ar+ laser and the 5145 A exciting
line. There was excellent agreement ($1.0 cml) between the
spectra obtained using the two light sources. Figure 11 is
the Raman spectrum of N-methyl nicotinamide iodide in DMSO
using the Kr+ laser and Figure 12 is the Raman spectrum of
Nemethyl nicotinamide iodide in D20 using the Ar+ laser.

Table 5 lists the values obtained for the carbonyl frequen-

cies Observed for the solutions and the solids.

30

Table 4. Melting point values for N-methyl nicotinaldehyde
iodide and N-methyl 3-acetyl pyridine iodide in

 

 

C.
Compound Thls Literature values
work

Nemethyl
nicotinalde- 174—175 172.5-173.5 Boger, Black, San Pietro
hyde iodide 173 Ginsburg, Wilson

173-175 Ellin, Kondritzer

174 Pannizon

N-methyl
B-aIEFK: 163—165 163.5 Boger, Black, San Pietro
Pygid: 163-164 Ginsburg, Wilson
10 162 Pfleiderer, Sann, Stock

160-163 Akagi, Paretsky

 

 

 

P. Boger, C. Black, and A. San Pietro, Biochem., 6, 80
(1967)

S. Ginsburg and I. Wilson, J. Am. Chem. Soc., 19, 481
(1957)

R. Ellin and A. Kondritzer, Anal. Chem., 31, 200 (1959)
L. Pannizon, Helv. Chim., Acta 24, 24E (1941)

G. Pfleiderer, E. Sann, and A. Stock, Chem. Ber., 23,
3083 (1960)

J. Akagi and D. Paretsky, J. Org. Chem., 24, 152 (1959)

31

1696

Wavenumber in cml

 

808T—
68LI—
OLLT -
ISLI~
ZELIH
SILI—
17691—
9191:-
9991-
L£9I

8I9I-
665T—
0851-«
I9SI—

Figure 11. Raman spectrum of N-methyl nicotinamide iodide
in dimethylsulfoxide.

1667

32

 

 

 

 

. -l
Wavenumber in cm

 

1746-1
1715 -
1683-
1652-:

Figure 12. Raman spectrum
iodide in D20.

1620-
1588-
1556..

of N—methyl nicotinamide

33

Table 5. Position of the carbonyl frequency for N-methyl
nicotinamide iodide, Nrmethyl 3vacetyl pyridine
iodide, and N—methyl nicotinaldehyde iodide.

 

 

 

IODIDE SALTS
Solvent N-methyl N-methyl N-methyl
nicotinamide nicotinaldehyde 3-acetyl
pyridine
D20 1667 (1637) 1720 (1705) 1704 (1684)
DMSO 1696 (1686) 1715 (1702) 1702 (1688)
pure 1680 1699 1702
(solid)

 

 

 

The values in parenthesis are the values of the non-
methylated compounds from Table 1.

Error limit i 1.0 cml for all carbonyl positions.

34

Results and Conclusion

 

Upon inspection of the data in Table 5, one finds
there has been a slight increase in the carbonyl frequency
(on the order of 10-20 cml) when the values of the methyl-
ated compounds are compared with those for the nonmethyl-
ated derivatives. An increase in the carbonyl frequency
strongly suggests that the electrons of the carbonyl group
have become more localized, resulting in an increased dou—
ble bond character. Thus, there is no indication that the
presence of a plus charge on the pyridine ring system re-
sults in further delocalization of electrons from the car—
bonyl group into the ring.

Of the solvents used in the previous study (Table
1), only deuterium oxide and dimethylsulfoxide would dis-
solve a sufficient concentration for Raman spectra of rea-
sonable intensity. Therefore, a solvent study as extensive
as the one performed on the nonmethylated derivatives could
not be completed.

As shown in Table 5, N—methyl nicotinaldehyde and
N—methyl 3-acetyl pyridine iodide, like the nonmethylated
compounds, shows only minor differences between the car-
bonyl frequencies in the different solvents. Thus, if an
extensive solvent study could have been done, the results
obtained would probably have been very similar to the data

obtained for the nonmethylated derivatives.

35

As with nicotinamide, however, N-methyl nicotin-
amide iodide shows an anomalous effect on the carbonyl fre-
quency with the presence of a hydrogen bonding solvent.

The difference between the DMSO and the D20 values is 29
cml as compared with 49 cml for nicotinamide. One can,
again extend the same arguments that were advanced for the
anomaly of nicotinamide in solution to explain the anomaly
of N-methyl nicotinamide in solution. The electronic dis—
tribution of the carbonyl group has been changed due to a
resonance effect between the carbonyl group and the adja-

cent amino group and this effect is facilitated by the

presence of a hydrogen bonding solvent.

CHAPTER III

ANALYSIS OF THE CARBONYL
FREQUENCY OF NAD

Introduction

 

In their circular dichroism study, Miles and Urryl

noted that a reciprocal behavior was observed for both
beta NAD+ and NADH as the temperature was increased above
room temperature. They concluded that at room temperature
NAD exists in a stacked conformation and the reciprocal be-
havior that is observed as the temperature is increased to
65°C. is due to the molecule changing conformation and
going to an open form, that is, no stacking between the
adenine and the nicotinamide moieties.

With this observation in mind, studies were per-
formed on beta NAD+, NADH, the 3-a1dehyde analog of NAD+,
NMNH, and N-methyl nicotinaldehyde iodide to determine the
effects of temperature on the carbonyl frequency of these
compounds. NMNH is the abbreviation for the reduced form
of nicotinamide mononucleotide and Figure 13 shows the
structure of this compound. In addition, the 3-acetyl an-
alog of NADH was examined at room temperature. Ultraviolet

spectra were also obtained in several cases in order to

36

37

 

Figure 13. Structure of NMNH.

38

investigate possible correlation between the electronic

and vibrational spectra.

Experimental

 

Beta NAD+, NADH, NMNH, and the 3-a1dehyde analog of
NAD+ were all obtained from the Sigma Company and the 3-
acetyl analog of NADH was obtained from P-L Biochemicals,
Inc.. All were used without further purification. The
solvents used were deuterium oxide (Columbia Organic Chem-
icals) and dimethylsulfoxide (J. T. Baker). The dimethyl-
sulfoxide was purified as described previously. The N-
methyl nicotinaldehyde iodide was that prepared in the pre-
vious chapter (see Chapter II, Experimental).

A Coleman 124 double beam Spectrophotometer and
quartz cells were used to take UV spectra. The following
compounds were inspected by UV: 0.1mM NADH in DMSO and
H20; 0.2mM N-methyl nicotinamide iodide, N-methyl nicotin-
aldehyde, and N—methyl 3-acetyl pyridine iodide in H20;

0.1mM NMNH in H O; and a 0.1mM solution of the 3-acetyl an—

2

alog of NADH in H O and DMSO. The values obtained are

2
listed in Table 6.

The Raman and IR spectrometers and cells used were
those described in the first chapter (see Chapter I, EXper-
imental).

Temperature studies were carried out by placing the

sample capillary in an unsilvered Dewar cell, as depicted

h
u?
-1

 

 

39

Table 6. Ultraviolet spectra of selected compounds.

 

 

 

 

Pyridine -Dihydro A in nm
Derivatives Derivatives max.
Solvent This
R2 R3 R4 R5 System Work Literaturezo'21
NH2 CH3 H20 264 ---
CH3 CH3 H20 265 ---
H CH3 H20 262 ---
NH2 ribose 5 H20 --- 263
phosphate
NH2 M H20 338 338
NH2 M DMSO 332 ---
CH3 M H20 362 363
CH3 M DMSO 354 ——-
H M H20 ~-- 358
NH2 CH2C6H5 CH3OH --- 352
CH3 CH2C6H5 CH3OH --- 371
NH2 ribose 5 H20 337 ---
phosphate

 

 

 

 

Error limit: i 1.5 nm

Basic Structures:

Pyridine Derivatives Dihydro Derivatives
'6 c'i
\\\ \
R | 1 \R
2 4
-+//
N
[1? l
3 R

M: Structure from Figure l, excluding the R group

 

40

in Figure 14. Nitrogen gas was passed through coils wrapped
in heating tape and placed in an auxillary Dewar. The flow
rate of the nitrogen gas, which could be accurately con-
trolled, determined the temperature. Using a Leeds and
Northrup millivolt potentiometer, the temperature was mea-

sured in the sample chamber by a Copper-Constantan thermo-

‘VD

couple placed in a mercury—filled tube adjacent to the sam-

ple.

Jrfiiifik‘. m .‘n‘fu

Raman spectra were taken of NAD+, NADH, NMNH, the

"out
a.

‘21,

aldehyde analog of NAD+, and N-methyl nicotinaldehyde io-
dide at room temperature (19°C.), and at 70°C. The 3-
acetyl analog of NADH was examined at room temperature
(21°C.). The NADH, NMNH, and the 3—acetyl analog of NADH
solutions were scanned using the Kr+ laser and the 6471 A
exciting line and the others were inspected using the Ar+
laser and the 5145 A exciting line. Solutions of 0.25M
were prepared in D20. NADH and the 3—acetyl analog of NADH
were also scanned in DMSO solutions. The solubility prop—
erties of the compounds studied prevented any extensive ex-
amination of the solvent effect upon the carbonyl frequen—
cies. Table 7 lists the carbonyl frequency values obtained
in the temperature studies.

Figure 15 is the Raman spectrum of NMNH in D20 at
19°C. and was obtained using the Kr+ laser. Figure 16 is

the IR spectrum of a 0.1M NAD+ solution in D 0. Figure 17

2

is the Raman spectrum of NAD+ in D O at 19°C. using the Ar+

2

laser.

 

41

Copper-constantan
thermocouple * Laser Beam
N exit H

*

       

to

Mo chromator

      
 
 

N2 entrance

  

Side view Raman sample
Cork cell

 

 

 

 

 

 

 

 

 

 

 

Figure 14. Raman temperature control uni 22.

42

Table 7. Effect of temperature on the carbonyl
frequency in the Raman spectra of NAD
and similar molecules.

 

 

 

Carbonyl Frequency (cml)
Compound
Temperature 0

19 C. 70 C.
NAD+ in D20 1667 1667
NAD+ in D20 1666 (IR)
3-Aldehyde analog
of NAD+ in D20 1717 1717
N-methyl nicotin-
aldehyde iodide
in D20 1720 1720
NMNH in D20 1689 1689
NADH in H20 1689 1689
NADH in D20 1689 1689
NADH in DMSO 1689 1689
3-Acetyl analog
of NADH in D20 1679*
3—Acety1 analog
of NADH in DMSO 1679*

 

 

Error limit:

rl.o oil for all carbonyl
positions.

0
*Temperature was 21 C.

IR infrared value

43

1689

 

 

-1

Wavenumber in cm

 

Hmmn

Hme

Hmuw

HmmH

quo

Hmmo

quq

qum

anm

Hump

quw

Hmom

Raman spectrum of NMNH in D20.

Figure 15.

44

__1800
"l720
1-1640
“1560
"1480
‘1400

 

. -l
Wavenumber in cm

 

 

 

 

 

 

 

 

1666

 

Figure 16. IR spectrum of NAD+ in D20.

Figure 17.

45

 

 

 

 

 

 

 

 

 

1667

Wavenumber in cml
1 I F l l T1 l I l
\ommmoaoxovm
VHmmNmLfiNO‘
Fl‘kDKOkDLnl-nlnq'
.4 H —a pa .4 H F4 .4 H
Raman spectrum of NAD+ in D O.

2

46

Results and Conclusion

 

Intramolecular hydrogen bonding is an internal ef-
fect and persists at very low concentrations, but the ex-
tent of intramolecular bonding would be expected to be

temperature dependent.13

If an intramolecular hydrogen
bond existed between the adenine amino group and the car—
bonyl group of the nicotinamide, it should be reflected in
a lowering of the carbonyl frequency.12 With an increase
in temperature, which has been shown1 to disrupt the inter-
actions between the rings, the perturbation in the carbonyl
frequency would be removed. As shown in Table 7, there is
no change in the carbonyl frequency of NAD+, NADH, and the
3valdehyde analog of NAD+, with the change in temperature.
This lack of a temperature dependence of the carbonyl fre-
quency in the compounds considered may be interpreted as
evidence against the existence of an intramolecular hydro-
gen bond between the amino adenine group and the nicotin-
amide carbonyl group of NAD.

There is also excellent agreement between the car-
bonyl frequencies observed for NAD+ and the aldehyde analog
of NAD+ (Table 7) and the values found for the correspond-
ing N-methyl derivatives (Table 5); this provides further
support for the view that the N-methyl compounds represent
useful models for the more complex pyridine nucleotides.
Furthermore, the agreement between the position and the ef-

fect of solvent on the carbonyl frequency of NAD+ and that

47

of the model compound, N—methyl nicotinamide iodide, indi-
cates that the resonance interaction proposed to exist in
the model compound (see Chapter II) also exists in NAD+.

When one scans Table 6 for the oxidized methylated
compounds that were considered in Chapter II, one finds
that the values obtained are very similar to NMN. Thus,
the electronic transitions in the ring system of the oxi-
dized compounds are insensitive to changes in the substit-
uents on the ring nitrogen or the three position of the
ring. This again indicates that the N-methyl derivatives
serve as adequate model compounds for the NAD oxidized an-
alogs. Also, the suggested resonance interaction between
the carbonyl group and the adjacent amino group apparently
does not affect the electronic transitions of the ring.

In contrast, the UV data indicate that the elec-
tronic transitions in the dihydro compounds a£§_appreciably
affected by the nature of the substituents on the nitrogen
of the ring and in the three position. The data in Table
6 show that when the substituent in the three position is
held constant, there is a difference in the absorption max-
imum dependent on the substituent in the one position of
the dihydro ring. When one keeps the substituent in the
one position constant, one finds that changing the substit-
uent in the three position also results in major changes.
Thus, it is clear that, in contrast to the oxidized com-

pounds, there must be extensive electronic interactions

48

between the group at the three position (which includes the
carbonyl of direct interest in the present study) of the
dihydro pyridine ring and the rest of the molecule. The
nature of these interactions in the dihydro compounds can
only be elucidated by a more extensive study.

Sarma and Kaplan23

stated at a symposium on "Pyri-
dine Nucleotide-Dependent Dehydrogenases" that: "It is
difficult to generalize why the different analogs substi-
tuted in the three position of the pyridine moiety react so
differently with the various pyridine nucleotide dehydroge-
nases. The variation is certainly not attributable to dif-
ference in the stacking of the bases. The possibility
exists that different enzymes can distinguish changes in

the side chain at the three position." Thus, the electronic
distribution of the amide group of NAD+ and NADH could be

of major importance in determining interaction at the bind-
ing sites for the coenzymes. The present investigation pro-

vides considerable evidence in support of the view that, in

aqueous solutions, the three position 3 group of
/c - NH2

NAD+ is not a "typical" amide, but rather is more adequately
represented by the structure:

oJ‘
\ 1+

 

49

Furthermore, this structure does not appear to make a sig-
nificant contribution to the electronic structure of NADH.
Thus, changes in oxidation state of the ring affect not
only the ring itself, but also the electronic properties of
the three substituent. These differences are likely to be
of great significance in determining enzyme-coenzyme inter-

actions.

LI ST OF REFERENCES

11.

12.

13.

14.

LIST OF REFERENCES
D. W. Miles and D. W. Urry, J. Biol. Chem., 243, 4181
(1968).

T. G. Scott, R. D. Spencer, N. J. Leonard, and G. Weber,
J. Am. Chem. Soc., 22, 687 (1970).

J. A. Secrist and N. J. Leonard, J. Am. Chem. Soc., 24,
1702 (1972).

R. H. Sarma, P. Dannies, and N. 0. Kaplan, Biochem., l,
4359 (1968).

R. H. Sarma and N. 0. Kaplan, Biochem. Biophy. Res.
Commun., 24, 780 (1969).

R. H. Sarma and N. 0. Kaplan, Biochem., g, 539 (1970).
R. H. Sarma and N. 0. Kaplan, Biochem., g, 557 (1970).

D. W. Miles and D. W. Urry, J. Phys. Chem., 12, 4448
(1967).

J. Jacobus, Biochem., 24, 161 (1971).

M. J. Adams, A. McPherson, Jr., M. G. Rossman, R. W.
Schevitz, and A. J. Wonacott, J. Mol. Biol., 52,
31 (1970).

S. N. Vinogradov and R. H. Linnell, Hydrogen Bonding,
Van Nostrand Reinhold Co., Cincinnati, 1971, p. 18.

 

See reference 11, p. 75.

R. M. Silverstein and G. C. Bassler, Spectrometric Iden-
tification of Organic Compounds, John Wiley and
Sons, Inc., New York, 1967, p. 87.

 

R. E. Miller, K. L. Treuil, R. R. Getty, and G. E.
Leroi, Contract NOOOl4-68-A-0109-0003; NR 014-203,
TR No. 4(29), (1968). DDC Aceession No. 669797.

50

15.

16.

17.

18.

19.

20.

21.

22.

23.

51

A. L. McClellan, Table of Experimental Dipole Moments,
W. H. Freeman and Co., New York, 1963.

 

Stanley W. Jacob, E. E. Rosenbaum, D. C. Wood,
Dimethyl Sulfoxide, Marcel Dekker, Inc., New York,
1971, p. 5.

 

Robert C. Weast, Ed. Handbook of Chemistry and Physics,
Chemical Rubber Company, Cleveland, 52, 1972-1973,
E 43’460

See reference 13, p. 68.

P. Boger, C. C. Black, and A. San Pietro, Biochem., 6,
80 (1967).

U. Eisner and J. Kuthan, Chem. Rev., 12, l (1972).

J. M. Siegel, G. A. Montgomery, and R. M. Bock, Arch.
Biochem. Biophy., 22, 288 (1959).

Douglas A. Hatzenbuhler, PhD Thesis, Michigan State
University, Department of Chemistry, 1970, p. 46.

H. Sund, Ed., Pyridine Nucleotide-Dependent Deh dro e-
nases, Springer-Verlag Berlin, New York, 197%,

p. 44-45.