Essentialsin "

Po_er Pr¢_enta¢ion

5th Symposium of the Protein Society - June, 1991Poster # T38

Adsorptive Protein and Peptide SequenceAnalysis on Chemically-Modified

Polyvinylidine Difluoride Membranes

J.D. Dixon, D.W. Andrews & J.M.Coull _

Millipore Corporation, MilliGen/Biosearch Division, <C:186 Middlesex Turnpike, Burlington, MA. 01803 U.S.A. '

J

PaR Waters Chromatography Division 7

MILLI E 34 Maple StreetMilford, MA 01757508 478-2000

ABSTRACT:

Polyvinylidene difluoride (PVDF) membranes arecommonly utilized for adsorptive sequence analysis ofpolypeptides. Sequence analysis of short and/orhydrophobic peptides is improved by the addition ofpolymeric cationic carriers such as polybreneTM orpolyethyleneimine. Unfortunately, these carriers requiretime consuming precycling before sample application andcontribute to increased background peaks in the HPLCanalysis. We have utilized a chemically-modified-membrane which does not require precycling andfacilitates the sequunce analysis of small peptides. Alsoimproved initial step yield and repetitive sequencingyields from protein or peptide samples applied or blottedto PVDF are obtained when a chemically-modified-membrane is added to the ProSequencer TM reactionchamber as a trapping membrane.

INTRODUCTION:

In an adsorptive sequencing mode glass supportsoften have polymeric cationic carriers such as polybreneTM(Tarr et aL, 19"8) added to improve the retention of smallpeptides on _ese supports (Hewick et al., 1981). Glass-fiber sheets have also been modified to allow the covalent

attachment of polypeptide chains (Abersold et al., 1986).There has been a shift in the last few years from glass-based solid supports to PVDF membranes (Matsudaira1987). Membrane supports offer the advantages ofmechanical and chemical stability to a wide variety ofconditions and they are easily merged with other analyticaltechniques such as sodium dodecylsulfate-polyacrylamidegel electrophoresis. Functionalized PVDF based supportshave also been developed to allow the covalent attachmentof proteins and peptides (Coull et aL, 1989; Coull et al.,1991).

We report here the use of an quarternary aminemodified PVDF membrane (Immobilon-NTM) as an effectivesupport for the adsorptive sequence analysis of smallpeptides.

MATERIALS & MET: £)DS:

The Immobilon-N and -P membranes were obtainedfrom Millipore in 25 x 25 cm sheets. Batches of 8 mm diskswere prepared from the sheets and stored at -20 oC. A6625 ProSequencerTM as described in poster T63 was usedto evaluate the various membrane supports.

Two test peptides were synthesized at the 0.1 mmolscale on a methylbenzhydrylamine polystyrene-basedsupport PACTM using an ExcellTM peptide synthesizer.Peptides were synthesized using 9-fluorenyl-methoxycarbonyl (Fmoc) alpha-amino protected aminoacids and benzotriazolyloxy-trisdimethyl-aminophosphonium hexafluorophosphate (BOP)/1-hydroxybenzotriazole (HOBt) activation. The acylationtimes used for each amino acid addition were determined

by an expert system. All of the peptide synthesis andprotein sequencing chemicals were supplied byMilliGen/Biosearch.

Stock solutions of the test peptides were prepared in25% (v/v) acetonitrile/ Milli-Q water (Fluka ChemicalCorporation, Ronkonkoma, NY) An aliquot of 10 uL (200pmol) was applied to the center of the sample disks andallowed to air dry before use. The horse heart myoglobinsample was collected from a prototype capillaryelectrophoresis system.

RESULTS & DISCUSSION:

The membrane supports were evaluated in terms ofthe amount and number of background peaks observed oneach type of blank support (Figures 1-3). The Immobilon-Psupports provided chromatograms with the smallestnumber of interfering background peaks and at lowerlevels than the Immobilon-N membranes (Figure 4). Therewas a slight advantage in using the modified Immobilon-Nmembrane over the Immobilon-N, as lower initialbackground peaks were observed.

Small peptides less than 10 residues could besequenced on either of the Immobilon-N membranes to theC-terminal amino acid (Figure 5). Useful sequenceinformation wa_ obtain on Immobilon-P, although the C-terminal residue was not identified. Factors such as the

hydrophobicity of the sample and composition of thepeptide will affect the peptide removal rate from themembrane during the sequencing process.

The addition of a backing or trapping membrane ofImmobilon-N resulted in the determination of the C-

terminal amino acid from peptide 1 applied to anImmobilon-P support (Figure 6). This is significant aspolypeptides are easily detected on Immobilon-Pmembranes by various visualization techniques. Thesample could then be retained by simply placing aImmobilon-N membrane underneath the Immobilon-P

support. The low background levels of the Immobilon-Pmembrane aided the sequence analysis of Iow-picomoleamounts of sample (Figure 7).

CONCLUSIONS:

• Immobilon-P membranes yield a cleaner backgroundthan the Immobilon-N based supports.

• Immobilon-N or the modified version retains smallpeptides better than Immobilon-P.

• Immobilon-N can be used as a backing membrane toretain sample which is washed off the primarymembrane.

• Immobilon-N and -P can be used for adsorptivesequence analysis without precycling the membrane.

ACKNOWLEDGMENTS:

The authors thank Henry Cheng, Martin Fuchs and BillCarson for the horse heart myoglobin sample. Theassistance of Brian Gildea in the preparation of this posteris gratefully acknowledged.

REFERENCES:

Aebersold, R.H., Teplow, D.B., Hood, L.E., & Kent, S.B.H. (1986)Electroblotting onto activated glass; high efficiencypreparation of proteins from analytical sodiumdodecylsulfate-polyacrylamide gels for direct sequenceanalysis. J. Biol. Chem. 261"4229-4238.

Coull, J.M., Dixon, J.D., Laursen, R.A., KSster, H., & Pappin, D.(1989) in: Methods in protein sequenceanalysis.(Wittmann-Liebold, B., Ed.) "Development of membranesupports for the solid-phase sequence analysis of proteinsand peptides", pp. 69-78, Springer-Verlag, Berlin.

Coull, J.M., Pappin, D.J.C., Mark, J., Aebersold, R., & KSster, H.(1991) Functionalized membrane supports for covalentprotein microsequence analysis. Anal. Biochem. 194:110-120.

Hewick, R.M., Hunkapiller, M.W., Hood, L.E., & Dreyer, W.J.(1981) A gas-liquid-solid-phase peptide and proteinsequenator. J. Biol. Chem. 256:7990-7997.

Matsudaira, P. (1987) Sequence from picomole-quantitiesof proteins electroblotted onto polyvinylidene difuoridemembranes. J. Biol. Chem. 262:10035-10038.

Tarr, G.E., Beecher, J.F., Bell, M., & McKean, D.J. (1978)Polyquarternary amines prevent peptide loss fromsequencers. Anal. Biochem. 84:622-627.

Blank Immobilon-P Membrane

20000

15000

¢_ ./_z'_'IZ_ _.,._z_ I,, 'm - Peak -1&q,=

= Peak -2x 10000

m -- Peak -3

o= Peak -4

5000

00 2 4 6 8 10 12

1A Cycle Number

Figure 1. Adsorptive mode blank cycles from Immobilon-P membrane. The amount of three interfering backgroundpeaks and diphenylthiourea (DPTU) are shown in Panel A.The values are compared in terms of peak height.Approximately 400 units is equivalent to 1 pmol ofphenylthiohydantoin (PTH) amino acid. The chromatogramsfrom the first and tenth cycles are shown in Panel B. The

PTH standard is at the 50 pmol level and the first cyclebelow the standard is traced in yellow.

e°

S .00

¢ 4.00

NI

0'_ 3. O0,

X

2.00 _

I ' ' ' I ' ' ' I ' ' ' 1 I IO. 40 0.60 0.80 I .00 t .20 1,40

x t0 ! Minutes

o

: _ 3.503

_ 3. O0

! 0

• . ql_

_'i X 2.50 2

o

I I [ [ [ [0.40 O.GO 0.80 1.00 I .20 t .40

1B x tO 1 .inutes

Blank Immobilon-N Membrane

20000

15000 -_

c_ ' --: Peak -1

x 10000 • Peak -2

-- Peak -3,.,..0

> • Peak -4

50000 t | _ I |

0 2 4 6 8 10 2

2A Cycle Number

Figure 2. Adsorptive mode blank cycles from Immobiion-N membrane. The amount of three interfering backgroundpeaks and DPTU are shown in Panel A. The values arecompared in terms of peak height. Approximately 400units is equivalent to 1 pmol of PTH amino acid. Thechromatograms from the first and tenth cycles are shownin Panel B. The PTH standard is at the 50 pmol level and thefirst cycle below the standard is traced in yellow.

4.00

3.OONI0

X

t

2.00

O. 40 0.60 0.80 1. O0 1.20 1.40

x t01 Minutes

o

m

3 .oo

Ji

2.00

4a

ii °_ x

_.! t .oo 1o : : 2

o

_.- , 4

j :0.00 , ,_ i ' i

O. 40 0.60 0.80 I .00 t .20 I . 40

2B x 1o I ninutes

Blank Modified Immobilon-N Membrane

20000

15000 H_ __ o Peak -1

x 10000 8 Peak -2

=o -- Peak -3o• o Peak -4

ooo. ,_' _ _. g. •0 I I I I I ='

0 2 4 6 8 10 12

Cycle Number

3A

Figure 3. Adsorptive mode blank cycles from a modifiedImmobilon-N membrane. The amount of three interferingbackground peaks and DPTU are shown in Panel A. Thevalues are compared in terms of peak height.Approximately 400 units is equivalent to 1 pmol of PTHamino acid (Panel A). The chromatograms from the firstand tenth cycles are shown in Panel B. The PTH standard isat the 50 pmol level and the first cycle below the standardis traced in yellow.

5.00

I

2.00

0.40 O.EO 0.80 1.00 1.20 1.40

x t01 -inutes

o

Q

_ 3.00-

3

2.50m

° ." f 2.00

ii_ °_,z m

_ x 2

ii_'" | .50 _ I 4

ooO. 40 O. 60 0.80 ".. O0 1.20 I. 40

3B x 101 -,nutes

Peak 1 Background Levels Peak 2 Background Levels6000 -,----------- - _ _ 6000

5000 5000

o;i 4000 _ 4000

o • PVDF-1 "" • PVDF-2

>¢ 3000 _ x 3000Imm-N -1 I_i Imm-N -2

2000 [] Modlmm-N-1 o 2000 [] Modlmm-N-2>

1000 1000

o 0

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

Cycle Number Cycle Numbers

Peak 3 Background Levels Peak 4 Background Levels20000 6000 °- ,

-, -, ,ooo11] Ix 10000 Imm-N-3 x 3000 Imm-N-4

[] Modlmm-N-3 _ [] Modlmm-N-4

I,°°I0

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

Cycle Number Cycle Number

Figure 4. Summary of background peaks from the threetypes of membranes. Where peak 1, aniline (co-elutes withPTH-Gly); peak 2, co-elutes with PTH-Trp; peak 3, DPTU;and oeak 4. co-elt_tp._with PTN-II_,

immoDilon-P Versus Immobilon-NPeptide 1

1000.

100

,-I-

ra = Imm- N-- ;30E • Imm-PQ.

10

10 2 4 6 8 10

5A Cycle Number

(1) +NH3-Asn Ser Ala Gin Arg Thr Gly Asn Glu CO0-

Immobilon-P Versus Immobilon-NPeptide 2

1000 .

100

-i-I,--n " Imm-N

0E • Imm-PD..

10

1 I . I _' I t I

5B o 2 , 6 8 loCycle Number

(2) +NH3-Tyr Met Lys Cys Val Pro Phe Tyr Glu COO-

Figure 5. Adsorptive mode sequence analyses of testpeptides 1 (Panel A) and 2 (Panel B) from Immobilon-P and-N membranes.

I

Immobilon-P Versus Immobilon-N

Backing Membranes Peptide 11ooo

lOO

.pI-n : Double Imm-N"6E • Double Imm-P=- 10

I I I I

0 2 4 6 8 0

Cycle Number

Figure 6. Adsorptive mode sequence analyses of testpeptide 1 from Immobilon-P with backing membranes ofImmobilon-P or -N.

"4,

HH Myogiobin on Immobilon P10

initial yield 1.9 prnolesRepe;itive yield 97.8%

13

.1

0 10 2O 30

CYCLE NUMBER

HORSE HEART MYOGLOBIN MW 16,950 Daltons

1 10 20GLSDGEWQQVLNVWGKVEAD

Figure 7. Horse heart myoglobin sequenced fromImmobilon-P membrane.