***immediate response required***cnc.cj.uc.pt/.../pdfs/rmn0607/nmr_biomed_celda_05.pdfearlyview® is...
TRANSCRIPT
TH E A T R I U M, S O U T H E R N G A T E, C H I C H E S T E R, W E S T S U S S E X P019 8SQ
***IMMEDIATE RESPONSE REQUIRED***
Your article may be published online via Wiley's EarlyView® service (http://www.interscience.wiley.com/) shortly after receipt of corrections. EarlyView® is Wiley's online publication of individual articles in full-text HTML and/or pdf format before release of the compiled print issue of the journal. Articles posted online in EarlyView® are peer-reviewed, copy-edited, author-corrected, and fully citable via the article DOI (for further information, visit www.doi.org). EarlyView® means you benefit from the best of two worlds - fast online availability as well as traditional, issue-based archiving.
Please follow these instructions to avoid delay of publication READ PROOFS CAREFULLY
• This will be your only chance to review these proofs. Please note that once your corrected article is posted online, it is considered legally published, and cannot be removed from the Web site for further corrections.
• Please note that the volume and page numbers shown on the proofs are for position only.
ANSWER ALL QUERIES ON PROOFS (Queries for you to answer are attached as the last page of your proof.) • List all corrections and send back via e-mail to the production contact as detailed in the covering e-mail, or mark all corrections directly
on the proofs and send the scanned copy via e-mail. Please do not send corrections by fax or in the post.
CHECK FIGURES AND TABLES CAREFULLY • Check size, numbering, and orientation of figures. • All images in the PDF are downsampled (reduced to lower resolution and file size) to facilitate Internet delivery. These images will appear
at higher resolution and sharpness in the printed article. • Review figure legends to ensure that they are complete. • Check all tables. Review layout, title, and footnotes.
COMPLETE CTA (if you have not already signed one)
• Please send a scanned copy with your proofs and post your completed original form to the address detailed in the covering e-mail. We cannot publish your paper until we receive the original signed form.
OFFPRINTS
• 25 complimentary offprints of your article will be dispatched on publication. Please ensure that the correspondence address on your
proofs is correct for despatch of the offprints. If your delivery address has changed, please inform the production contact for the journal - details in the covering e-mail. Please allow six weeks for delivery.
Additional reprint and journal issue purchases
• Additional paper reprints (minimum quantity 100 copies) are available on publication to contributors. Quotations may be requested from mailto:[email protected]. Orders for additional paper reprints may be placed in advance in order to ensure that they are fulfilled in a timely manner on publication of the article in question. Please note that offprints and reprints will be dispatched under separate cover.
• PDF files of individual articles may be purchased for personal use for $25 via Wiley’s Pay-Per-View service (see http://www3.interscience.wiley.com/aboutus/ppv-articleselect.html).
• Please note that regardless of the form in which they are acquired, reprints should not be resold, nor further disseminated in electronic or print form, nor deployed in part or in whole in any marketing, promotional or educational contexts without further discussion with Wiley. Permissions requests should be directed to mailto:[email protected]
• Lead authors are cordially invited to remind their co-authors that the reprint opportunities detailed above are also available to them.
• If you wish to purchase print copies of the issue in which your article appears, please contact our Journals Fulfilment Department mailto:[email protected] when you receive your complimentary offprints or when your article is published online in an issue. Please quote the Volume/Issue in which your article appears.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
NMR IN BIOMEDICINENMR Biomed. 2005;18:1–11Published online in Wiley InterScience (www.interscience.wiley.com). DOI:10.1002/nbm.1005
Metabolite identification in human liver needle biopsies byhigh-resolution magic angle spinning 1H NMR spectroscopy
Beatriz Martınez-Granados,1 Daniel Monleon,1 M. Carmen Martınez-Bisbal,1 Jose Manuel Rodrigo,2
Juan del Olmo,2 Paloma Lluch,2 Antonio Ferrandez,3 Luis Martı-Bonmatı4 and Bernardo Celda1*1Departmento de Quımica Fısica, University of Valencia, Valencia, Spain2Servicio de Hepatologıa, Departament of Medicine, Hospital Clınico Universitario, University of Valencia, Valencia, Spain3Departamento de Patologıa, Hospital Clınico Universitario, University of Valencia, Valencia, Spain4Servicio de Radiologıa, Hospital Universitario Dr. Peset, Valencia, Spain
Received 23 February 2005; Revised 10 September 2005; Accepted 20 September 2005
ABSTRACT: High-resolution magic angle spinning (HR-MAS) 1H NMR spectroscopy of intact human liver needle biopsies
has not been previously reported. HR-MAS NMR spectra collected on 17 specimens with tissue amounts between �0.5 and
12 mg showed very good spectral resolution and signal-to-noise ratios. One-dimensional 1H spectra revealed many intense
signals corresponding to cellular metabolites. In addition, some high molecular weight metabolites, such as glycogen and
mobile fatty acids, could be observed in some spectra. Resonance assignments for 22 metabolites were obtained by
combining the analysis of three different types of 1D 1H spectral editing, such as T2 filtering or the nuclear Overhauser effect
and 2D TOCSY and 13C-HSQC spectra. Biochemical stability of the liver tissue during up to 16 h of magic angle spinning at
277 K was studied. Biochemical trends corresponding to the different pathologies were observed, involving free fragments
of lipids among other metabolites. NMR signal intensity ratios can be useful for discrimination among non-pathological,
hepatitis C affected and cirrhotic liver tissues. Overall, this work demonstrates the applicability of HR-MAS NMR
spectroscopy to the biochemical characterization of needle biopsies of the human liver. Copyright # 2005 John Wiley &
Sons, Ltd.
KEYWORDS: high-resolution magic angle spinning; 1H NMR spectroscopy; liver needle biopsy; metabolite identification;
resonance assignment
INTRODUCTION
High-resolution magic angle spinning (HR-MAS) NMRspectroscopy is a powerful technique for the investigationof metabolites within different intact tissues (1–6). Theprecise determination of biochemical and metabolicprofiles in intact tissue promises to extend the possibi-lities of NMR as a medical diagnostic tool (1,7–10). For
non-solid or highly viscous liquids, HR-MAS NMRspectroscopy allows the reduction of most of the linebroadening associated to restricted molecular motion,chemical shift anisotropy, dipolar couplings and fieldinhomogeneity by high-rate spinning of the sample atthe magic angle �¼ 54.7� (11–13). The potential of HR-MAS applications to the study of biological tissues hasbeen widely demonstrated in the investigation of differentcellular alterations (14,15). In addition, HR-MAS NMRspectroscopy of intact tissues (ex vivo) provides furtheradvantages over traditional high-resolution liquid NMRof tissue extracts (in vitro) (7,16,17). High-resolutionNMR on extracts of excised tissues requires largeamounts of sample (> 0.25 g of tissue) (18). Likewise,the extraction process via protein precipitation methodsprevents the direct observation of membrane semi-mobilelipids. Moreover, extraction methods usually discrimi-nate metabolites on the basis of solubility in a particularsolvent. Although it has some minor limitations asso-ciated mainly with the spinning of the sample (spinningsidebands, spinning degradation effects and spinningtemperature gradients, among others), HR-MAS NMRis a non-destructive technique, which requires minimal
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
*Correspondence to: B. Celda, Departamento de Quımica Fısica,Universitat de Valencia, Edifici d’Investigacio, C/Dr. Moliner 50,Burjassot 46100, Valencia, Spain.E-mail: [email protected]/grant sponsors: Bruker Espana; Bruker Biospin France;Association for the Development of Research in NMR (ADIRM).Contract/grant sponsor: Generalitat Valenciana; contract/grantnumber: GV GRUPOS03/072.Contract/grant sponsor: European Integrated Project eTUMOUR;Contract/grant number: FP6-2002-LIFESCIHEALTH 503094.Contract/grant sponsor: Red de Grupo de Encefalopatıa del Institutode Salud Carlos III; Contract/grant number: 03/155.Contract/grant sponsor: Ministerio de Educacion, Cultura y Deporte,Spai.
Abbreviations used: CPMG, Carr–Purcell–Meiboom–Gill; HR-MAS,high-resolution magic angle spinning; HSQC, heteronuclear singlequantum coherence spectroscopy; NOE, nuclear Overhauser effect;NOESY, nuclear Overhauser effect spectroscopy; TOCSY, 1H–1H totalcorrelation spectroscopy.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
sample preparation and allows the observation of most ofthe tissue metabolites and dynamic interactions in anextremely reduced sample quantity.
The liver is the largest abdominal organ. It has a majorrole as storage center for many endogenous fuels of thebody. In addition, the liver acts as a major distributioncenter for exogenous energy supplies. Consequently, livermetabolism is mostly dedicated to macromolecule synth-esis and degradation. Moreover, it is the primary agentresponsible for the detoxification of metabolic residualproducts and most exogenous toxins. Obviously, the livermetabolic profile is highly dominated by the presence oflarge macromolecules with partially restrained mobility.This makes high-resolution NMR studies on liver tissuevery challenging. In addition, metabolic extraction ofliver tissue has large disadvantages as many of themacromolecules of interest are quickly degraded in thesample preparation process. As mentioned above, HR-MAS NMR spectroscopy has been proven to be anappropriate technique for the study of intact tissues. Inprevious studies, HR-MAS NMR spectroscopy has beenused for extensive intact rat liver tissue metabolic char-acterization (19–21). Resonances of most significantmetabolites, such as choline, free fatty acids, monomersand short chains of �- and �-glucose, glycogen andglycerol, were assigned and the applicability of thetechnique to the study of pathological conditions wasestablished. However, these studies were done after thekilling of the animal and the removal of the whole liver toobtain a minimal amount (typically from 10 to 40 mg) ofintact liver tissue for the spectroscopic studies. Needlebiopsy is a minimally invasive method to obtain samplesof intact liver tissue, being regularly applied in clinicalenvironments for different purposes.
The aim of this study was to demonstrate the applic-ability to clinical studies of 1H HR-MAS NMR spectro-scopy biochemical profile determination in intact livertissue obtained by needle biopsy. As part of a global
project for the study of hepatic encephalopathy by bothex vivo and in vivo NMR spectroscopy, we present HR-MAS spectra of liver tissue affected by two differentdiffuse pathologies: chronic hepatitis and cirrhosis. Themetabolic profiles of these two liver pathologies obtainedby HR-MAS may provide the basis for more accuratediagnosis. Comparison of these spectra with those col-lected for non-pathological liver tissue may also shedsome light on the biochemical changes underlying thedevelopment of these pathologies. To help in the accom-plishment of these objectives, in this work well-resolvedspectra of human liver tissue were obtained and assignedby both 1H and 13C HR-MAS NMR spectroscopy. Pre-liminary variations in metabolite composition related tothe different pathologies were also identified. 1H spectraobtained in a resection of a needle biopsy with only0.5 mg of tissue suggest applicability to clinical studies.
EXPERIMENTAL
Tissue collection procedure
This study involves the use of human liver biopsies.Informed consent was obtained from all patients. Seven-teen samples of liver tissue were obtained. One sample wasobtained during autopsy of a 76-year-old woman who haddied 24 h previously due to cardiac failure (sample 1). Twoother samples were resections obtained during abdominalsurgery. The remaining 14 were samples in excess fromliver needle biopsies. Table 1 provides information on thepatients and samples studied. All samples were cooledimmediately and kept at 193 K until use.
Sample preparation
Owing to the small size of the samples, all the material tobe in contact with the tissue was cooled to at least 277 K
Table 1. Human liver tissue samples studied in this work
No. Sample/weight (mg) Sex/age (years) Diagnosis Etiology
1 Autopsy/12 Female/76 Non-pathological liver Normal5 Biopsy/1.2 Female/66 Chronic hepatitis. Degree 2 Cryptogenetic7 Biopsy/7.5 Male/65 Cirrhosis VHCþ8 Biopsy/1.2 Female/50 Chronic hepatitis. Degree 0 Autoimmune9 Biopsy/2.9 Female/43 Chronic hepatitis. Degree 2 VHCþ
10 Biopsy/1.9 Male/40 Chronic hepatitis. Degree 1 VHCþ11 Biopsy/0.5 Male/42 Chronic hepatitis. Degree 2 VHCþ12 Biopsy/2.4 Male/56 Chronic hepatitis. Degree 1 Autoimmune13 Biopsy/– Male/32 Chronic hepatitis. Degree 2 VHCþ16 Surgery/9 Female/34 Non-pathological liver(metastasis) -17 Biopsy/– Female/57 Chronic hepatitis. Degree 2 VHCþ19 Surgery/– Male/72 Non pathological liver(metastasis) -21 Biopsy/– Male/66 Cirrhosis VHCþ22 Biopsy/– Male/41 Chronic hepatitis. Degree 2 VHCþ23 Biopsy/– Male/36 Chronic hepatitis. Degree 2 VHCþ26 Biopsy/– Male/46 Chronic hepatitis. Degree 3 VHCþ32 Biopsy/– Male/37 Cirrhosis VHCþ
2 B. MARTINEZ-GRANADOS ET AL.
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
to reduce tissue degradation during the sample prepara-tion process. Frozen samples were taken from a 193 Kfreezer and placed in a cryo-vial and in liquid N2 untilinsertion into a 4 mm outer diameter ZrO2 rotor. Thewhole operation takes no longer than 2 min. The pre-cooled rotor was filled with cooled D2O after sampleinsertion. Spherical inserts were used in all cases, limit-ing the rotor inner volume to 12ml. HR-MAS usingspherical small volume inserts provides improved resolu-tion and lineshapes compared with cylindrical inserts(19). D2O exceeding the 12ml volume was removedbefore rotor sealing. Owing to the small amount of tissueand the possible degradation during the sample weigh-ingQ1 process, tissue samples were weighed only afterHR-MAS measurements.
1H HR-MAS NMR spectroscopy of intacthuman liver tissue
All spectra were recorded using a Bruker Avance DRX500 spectrometer (Bruker, Valencia, Spain) operating at a1H frequency of 500.13 MHz. The instrument wasequipped with a 4 mm triple resonance 1H/31P/13C HR-MAS probe. For all experiments, samples were spun at4000 Hz to keep the rotation sidebands out of the acquisi-tion window. Field homogeneity, required for adequateresonances, lineshape and spectral resolution, wasachieved by extensive coil-shimming using a 1D waterpresaturation experiment in interactive mode as control.The lactate doublet at 1.33 ppm was used for fieldhomogeneity shimming, as described elsewhere (22,23).The sample was maintained at 273 K, as measured with athermocouple internal system, using the cooling of theinlet gas pressures responsible for the sample spinning.Sample heating caused by rotor spinning, calculatedusing the chemical shift of the water resonance, wasonly 4 K (7). The number of transients varied from 400 to1024 for the 1D spectra and from 16 to 64 for the 2Dspectra, depending on the tissue sample amount. Waterpresaturation was used during the final 1 s of the recy-cling delay in all experiments for solvent signal suppres-sion. Experiments acquired always included 1Dpresaturation 1H spectra, a 1D NOESY spectrum, a 1DCarr–Purcell–Meiboom–Gill (CPMG) spectrum and a2D TOCSY spectrum. For the autopsy sample, an addi-tional 2D 13C-HSQC spectrum was acquired exclusivelyfor assignment purposes. Recycling delays were 1.5 s forall experiments, with the exception of the CPMG experi-ments where 2.5 s was used. Spectral widths were6000 Hz for 1H and 10 000 Hz for 13C to cover 12 and80 ppm, respectively. The TOCSY mixing time for homo-nuclear Hartman–Hahn transfer was 50 ms. Spin-lockwas achieved by a DIPSI2 pulse sequence train duringthe TOCSY mixing time. The NOESY mixing time forthe 1D NOESY was 100 ms. The echo time in the 1DCPMG experiment was 36 ms. Echo times for the 2D
CPMG experiment ranged from 0.4 to 800 ms. One 1Dpresaturation 1H spectrum was acquired between experi-ments for tissue degradation control at 15 min and 1, 3.5,10 and 24 h after sample insertion. Spectra processingincluded zero-filling to double number of points andmultiplication by a shifted bell cosine-square windowfunction. Two-dimensional spectra processing included astage of linear prediction to increase digital resolution.
Spectral analysis
A preliminary statistical analysis was performed overintensity ratios of cross peaks present in at least 90% ofthe TOCSY spectra for all samples. Although TOCSYsignals are modulated by many different factors besidesmetabolite concentration, they were selected for thisanalysis to overcome overlapping and simplify the ana-lysis. In addition, most abundant metabolites show sev-eral narrow cross peaks in TOCSY which can be used forintensity ratio analysis. Therefore, two-dimensional spec-tra were included in this analysis. Nevertheless, all dataused for the metabolite quantitative analysis were col-lected in a period not longer than 4 h after sampleinsertion. This minimized the possible effects of tissuedegradation in the quantitative analysis. Cross peakintensities were calculated using the SPARKY (24) soft-ware for TOCSY spectra for all samples except samples 5and 13, whose TOCSY spectra did not provide thenecessary signal-to-noise ratio. For cross peak integra-tion, first the corresponding signal was fitted to a Gaus-sian function and then integrated using built-in SPARKYfunctions. For those metabolites where several crosspeaks were detected in the spectra, those showing lowererrors in the Gaussian fitting were selected for the analy-sis. Pathological differences in metabolite ratios amongstthe three entities were tested using SPSS Version 12.
RESULTS AND DISCUSSION
Spectroscopic data
All NMR spectra showed narrow linewidths and adequatesignal-to-noise ratios with well-resolved spin–spin multi-plicities, as shown in Figs 1 and 2. The use of highlysymmetrical spherical cavity inserts provided high spec-tral resolution and sensitivity. However, owing to thesmall amount of tissue in some needle biopsies, hetero-nuclear experiments such as 13C-HSQC or 31P-HMQCdid not provide sufficient signal-to-noise ratios for com-plete resonance assignment or quantification. Therefore,the standard dataset for all samples, with the exception ofthe liver autopsy sample (sample 1), did not include thesetype of experiments. The larger tissue amount obtained inthe autopsy sample (sample 1) allowed the acquisition of
Q1
HR-MAS OF HUMAN LIVER NEEDLE BIOPSIES 3
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
a good 13C-HSQC spectrum. This spectrum was usedexclusively for assignment support purposes. For allsamples, different types of signal filtering were appliedto the 1D 1H NMR spectra shown here. Little or nofiltering was observed in the 1D 1H water presaturationNMR spectrum [Fig. 1(A)]. As a consequence, all par-tially mobile macromolecules could be seen, distortingthe baseline in the most peak-crowded regions (from 0.5to 4 ppm). In the liver, large molecules are widely present(cholesterol, lipids, fatty acids, etc.) and, although notobservable as sharp NMR signals, they greatly distort thespectral baseline. A short spin-echo delay in the CPMGexperiment filters most of the signals from moleculeswith extremely short T2 relaxation times. In contrast,molecules with longer T2 values do not relax completely
and their signals can be observed. A spectrum withimproved baseline together with the disappearance ofsome signals is the result of such filtering [Fig. 1(C)].Comparison between CPMG experiments recorded ondifferent samples can still provide useful biochemicalinformation. Finally, 1D 1H water presaturation NOESYspectra provided fairly complete metabolite informationtogether with a nearly flat baseline, as shown in Fig. 1(B).The nuclear Overhauser effect (NOE) mixing delaycauses a similar effect to the spin-echo on the CPMGpulse sequence with the difference that most medium-sized macromolecule resonances still experience a strongNOE effect and their signals are not completely sup-pressed. However, large macromolecule resonances, re-sponsible for most of the baseline distortion, are
Figure 1. Comparison of different 1D 1H HR-MAS spectra of 9mg of human liver tissue(sample 16) at 273K and spinning at 4000Hz acquired with different types of filtering:(A) water presaturation 1D 1H NMR spectrum, (B) 1D 1H–NOESY spectrum and (C) 1D-CPMG experiment with water presaturation during recycling delay
4 B. MARTINEZ-GRANADOS ET AL.
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
completely relaxed after the mixing delay. Water signalpresaturation together with the NOE effect provideadditional solvent signal suppression. Metabolite quanti-fication in the 1D NOESY spectrum can still be modu-lated by the intensity of the NOE effect and, therefore,cannot be recommended. Two-dimensional TOCSYspectra provided useful information for resonance assign-ment (Fig. 2). The second dimension allows one toidentify many signals otherwise overlapped in the 1Dexperiments.
Spectral assignment
Prior to the resonance assignment process, all spectrawere referenced using the relatively invariant CH3 ala-nine signal at 1.47 and 19.11 ppm in 1H and 13C,respectively. These values of chemical shift have beenvalidated in previous HR-MAS studies with samplescontaining DSSQ2 reference (7). This signal was presentin all the spectra. One-dimensional 1H spectra of the liverexhibit inherent assignment problems due to overlap ofsmall molecules signals with many broad signals frommacromolecules and lipids. However, the different typesof spectral filtering, together with the 2D spectra ex-plained in the previous section, allowed many cases ofsuch overlapping to be resolved. Furthermore, manymetabolite spin systems were identified by using pre-
viously reported data for rat liver (20). Resonance assign-ments reported for other human organs were alsoparticularly useful in the assignment of less abundantshared metabolites (7). Overall, the assignment of 80resonances led to the identification of 22 metabolites onthe spectra, as shown in Table 2.
In the 1D 1H spectra of intact human liver tissue,resonances of lactate and mobile fatty acids are veryintense. Spin–spin connectivities observed in the 2DTOCSY provided the basis for the identification ofdifferent proton resonances from mobile fatty acids.Some very clear splitting patterns are observed formany metabolites such as alanine at 1.47 ppm and pro-tons for �- and �-glucose monomers and short polymers,located at 5.24 and 4.65 ppm, respectively. Many alipha-tic proton resonances from amino acids and glycogenwere also identified in the 2D TOCSY spectra. Theglycogen H1C proton resonance at 5.40 ppm showsvery variable intensity in the different spectra. Thecreatine CH3 proton resonance at 3.02 ppm is observedfor intact liver tissue for the first time. This signal is veryimportant for metabolite quantification as it has beenreported to be relatively invariant. Other assigned reso-nances not previously observed in studies of liver tissuesinclude leucine, tyrosine, phenylalanine and glutamine.Glutamine has been reported to be a good marker forammonium metabolism alterations (25,26). The presenceof ammonium greatly affects many central nervous
Figure 2. Aliphatic proton region of TOCSY spectrum with 50ms mixing time for7.5mg of liver needle biopsy (sample 7) at 273K and spinning at 4000Hzshowing the assignment of the most significant metabolites
Q2
HR-MAS OF HUMAN LIVER NEEDLE BIOPSIES 5
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
system biochemical processes. The two-dimensional 13C-HSQC spectrum obtained on liver tissue sample 16provided further confirmation of most resonance assign-ments (Fig. 3).
The liver has a major role in the regulation of fatty acidsynthesis and degradation and in carbohydrate and pro-tein metabolisms. In addition, it stores glycogen, aminoacids and lipids. The restricted sample size did not allowus to identify and assign all proton resonances in all thespectra. However, resonances for all metabolites listed inTable 2 were observed in complete datasets (including 2DTOCSY and 2D CPMG) of needle biopsies weighing aslittle as 1 mg. Spectra used for resonance assignment ineach sample were collected in less than 4 h. The identi-fication and possible quantification of these metabolitesby HR-MAS may provide very interesting informationabout alterations of liver metabolism. The potential of thetechnique applied to needle biopsies for diagnosis pur-poses seems obvious.
Tissue degradation
Changes over time with respect to the 1D NMR spectra ofliver tissue are not dramatic, as shown in Fig. 4. Theobserved changes were located in very specific reso-
nances, indicating the evolution of particular metabolicdegradation mechanisms. The spectral resolution did notchange significantly over time, indicating a relative con-servation in field homogeneity during the data acquisitionprocess. In our experience, temperature control seemscritical for tissue stability during HR-MAS data acquisi-tion (7). However, some tissue degradation still takesplace, probably owing to the high spinning rates. Thevariation in the signal of lactate is a commonly usedmarker to detect tissue degradation by oxidation (27).Resonances at 1.33 and 4.12 ppm belonging to CH3 andCH protons of lactate experienced a slight increase in the1D 1H spectra over time during spinning, indicating somelevel of degradation. In all cases, the increase in thelactate signal intensity was less than 10% after 24 h ofspinning. The resonance at 4.12 ppm was specially usefulwhen the signal at 1.33 ppm overlapped other intenselipid resonances in the same region.
In order to identify the separate contributions oftemperature and spinning to the tissue degradation duringthe data acquisition time, a simple test was performed.The effect of the spinning on tissue degradation wastested using two pieces of the same liver autopsy sample.Both samples were prepared for measurement and placedin respective rotors. HR-MAS spectra were collected forboth samples. Then, the first sample was kept in a
Figure 3. Aliphatic 1H/13C region of 2D 13C-HSQC spectrum obtained from 9mg of humanliver tissue (sample 16) at 273K and spinning at 4000Hz showing the assignment of themost significant metabolites
6 B. MARTINEZ-GRANADOS ET AL.
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
Table 2. Resonance assignments of most significant metabolites in HR-MAS NMR spectra of human liver tissue at273K and spinning at 4000Hz
Metabolite Group 1H chemical shift (ppm) 13C chemical shift (ppm)
Glycogen C1H 5.42 –C3H 3.98 73.71C5/6H 3.94 –/61.83C4H 3.68 –C2H 3.61 –
Lipids I CH——CHCH2 5.33 –CH——CHCH2 2.80 26.12CH——CHCH2CH2 2.03 27.22
CH3CH2CH2 1.32 23.21CH2CH2CO 2.23 34.37Lipids II CH3CH2 0.91 14.34
CH3CH2CH2 1.17 –(n)(CH2)CH2CH2COOH 1.29/1.55/2.16 –/24.92/–
�-Glucose C1H 5.22 –C2H 3.54 72.33C3H 3.70 75.47C4H 4.29 –
C5H 4.08 69.1�-Glucose C1H 4.66 –
C2H 3.24 74.53C3/5H 3.48 78.47/–Half CH2C6H 3.76/3.98 –C4H 3.39 70.11
Phosphorylcholine PO3CH2 4.18 –NCH2 3.61 –
Choline Nþ(CH3)3 3.20 56.69CH2(NH) 3.52 69.98CH2(OH) 4.07 58.39
Glycerol CH(OH) 3.77 74.781,3-CH2(OH) 3.62/3.57 65.06
TMAO N(CH3)3 3.27 50.45Creatine N(CH3) 3.03 39.71
CH2 3.93 56.72Lactate CH 4.12 –
CH3 1.33 22.66Acetate CH3 1.93 25.74Glutamate �CH 3.75 57.41
�CH2 2.34 36.14�CH2� 2.07 29.61�CH2m 2.04 –
Glutamine �CH 3.77 56.24�CH2 2.44 33.56�CH2 2.12 16.53
Lysine �CH 3.76 56.21�CH2 1.90 26.33�CH2 1.46 24.15�CH2 1.71 29.19"CH2 3.01 41.93
Leucine �CH 3.74 56.21�CH2 1.70 42.32�CH 1.71 27.10�CH3 0.95 23.61
Isoleucine �CH 3.61 62.01�CH 1.99 38.77�CH2m 1.26 27.11�CH3 1.00 17.29
Valine �CH 2.28 31.69�CH3 0.98/1.04 19.42/21.10Tyrosine CH3,5 6.88 –
CH2,6 7.18 –Phenylalanine CH2,6 7.31 –
CH4 7.38 –Alanine �CH 3.77 53.10
�CH3 1.47 19.11Dimethylglycine N(CH3)2 2.93 37.33
CH2 3.57 –
HR-MAS OF HUMAN LIVER NEEDLE BIOPSIES 7
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
refrigerator at 4�C for 24 h, whereas the second samplewas spun at 4�C for the same period of time, and HR-MAS spectra were collected again for both samples.Although some minor changes were observed for bothsamples, spectra collected for the sample that was spun
for 24 h displayed significantly larger changes in theintensity of the lactate signal (around 9% intensitychange compared with 2% observed for the sample thatwas not spun).
Interestingly, the intensity of the glycogen signalsshowed a substantial increase over time on the spinningrotor. Glycogen has an essential role in glucose home-ostasis. Alterations in glycogen metabolism take place inmany diseases, including some types of diabetes. Thelarge molecular size of glycogen makes the observationof its 1H resonance signals very difficult. In the HR-MASspectra of human liver tissue at low temperature, theresolution of the signal of the H1C glycogen protonseems to improve over time, suggesting an increase inits T2 relaxation time. This is partially confirmed by theobservation of glycogen signals in the 2D CPMG experi-ments performed on samples after 3 h of rotor spinning.This increase in T2 relaxation time is responsible for theobservation of this signal and its increment over time inthe HR-MAS NMR 1D 1H spectra. Longer T2 relaxationtimes may be caused by some macromolecule degrada-tion of glycogen to smaller fragments and ultimately toglucose units. Smaller molecules suffer slower transver-sal relaxation, giving rise to narrower signals. PreviousHR-MAS studies performed on rat liver tissue suggestthis as the most probable reason for the presence ofglycogen signals in 1H 1D NMR spectra (20). In this
Figure 4. Aliphatic proton region of 1H water presaturation1D spectra of human liver needle biopsy of 7.5mg (sample7) at 273K and spinning at 4000Hz after (A) 15min and (B)1, (C) 3.5, (D) 5, (E) 10 and (F) 24 h in the MAS rotor
Table 3. Most discriminative TOCSY cross peaks (in parentheses) intensity ratios for different metabolites(numbers of patients used for the analysis were three for non-pathological, eight for hepatitis C-affected andthree for cirrhotic liver, except where numbers in parentheses are given)
Metabolite ratio/classification Mean� SD
Alanine (�CH/�CH3)/lipids I (CH——CHCH2CH2/CH3CH2CH2)Non-pathological 0.13� 0.06Hepatitis 0.09� 0.03Cirrhosis 0.01� 0.00Lipids II (CH3CH2/(n)(CH2)CH2CH2COOH)/lipids I (CH——CHCH2CH2/CH3CH2CH2)Non pathological 0.19� 0.06Hepatitis 0.21� 0.03Cirrhosis 0.32� 0.11Leucine (�CH/�CH2)/lipids I (CH——CHCH2CH2/CH3CH2CH2)Non pathological 0.08� 0.09Hepatitis 0.07� 0.02Cirrhosis 0.01� 0.00�-Glucose (C2H/C1H)/lipids I (CH——CH—CH2—CH2—/CH3—CH2—CH2—)Non-pathological 0.52� 0.28Hepatitis 0.83� 0.21Cirrhosis 0.32� 0.15�-Glucose (C3H/C4H)/�-glucose (C2H/C1H)Non-pathological 1.03� 0.21Hepatitis 0.72� 0.31Cirrhosis 0.43� 0.11Leucine (�CH/�CH2)/lysine (�CH/�CH2)Non-pathological (2) 0.48� 0.42Hepatitis (7) 0.40� 0.24Cirrhosis 0.60� 0.08Alanine (�CH/�CH3)/lysine (�CH/�CH2)Non-pathological 2.93� 0.21Hepatitis (7) 0.63� 0.38Cirrhosis (2) 0.40� 0.35
8 B. MARTINEZ-GRANADOS ET AL.
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
study, this hypothesis is supported by a parallel increasein �- and �-glucose resonances in the 1D spectra overtime and the relatively long T2 relaxation times detectedby the 2D 1H CPMG experiment. Analysis of NMRsignals from glycogen during the first minutes of rotorspinning may be useful in the investigation of glycogenmetabolism. However, as only truncated glycogen mole-cules are observable by 1H NMR, some estimation of theproportion of degraded glycogen with respect to the totalglycogen amount may be necessary.
Biochemical comparison
Visual preliminary analysis of the spectra recorded for all17 patients show some interesting trends. Major metabo-lites in liver biochemistry include lipids, lactate, glucoseand glycogen. Based on the biochemical profile of someof these metabolites, preliminary correlations betweenspectroscopic data and the different pathologies shown inTable 3 have been found. Given the wide range of tissueweight, quantitative comparisons would need to be cor-rected for tissue weight. However, this type of quantifica-tion is connected with several sources of error (forexample, sample weight may include some blood orliquid). For our quantitative analysis we preferred towork on intensity ratios, as indicated below. Neverthe-less, some general trends can be immediately detected byvisual examination of the spectral profiles. In general, therelative signal intensities of mobile fatty acids, lipids andlactate increase from autopsy of non-pathological livertissue to hepatitis and cirrhosis (Fig. 5). In the case ofcirrhosis, the mobile fatty acid signals are very intensecompared with the rest of the metabolites. Glycogen alsodisplay a significant increase in intensity in cirrhosissamples with respect to hepatitis-infected liver. Althoughother general trends can be detected in the spectra, thewide range of sample sizes requires some type of meta-bolite quantification for accurate results.
Absolute metabolite quantification in HR-MAS spec-troscopy is a very challenging task. The usual strategy forobtaining classifiers and statistical variables for HR-MASspectra is to use the intensity ratios instead of absoluteintensities (28,29) (typically, ratios to creatine in brainspectra, as it is a relatively invariant metabolite). This isthe approach chosen here to analyze the results better.However, in liver tissue NMR spectra, no invariantmetabolite has been reported to date. For that reason,we calculated intensity ratios among cross peaks com-mon to at least 90% of the TOCSY spectra (to avoidoverlapping problems) and performed a preliminarystatistical analysis. Table 3 shows the most representativemetabolite cross peak intensity ratios selected for dis-crimination among the three different clinical situationsand their corresponding statistics. In general, the mostdiscriminative intensity ratios seem to include lipids,especially to differentiate between cirrhosis and no cir-
rhosis [Fig. 6(A) and (B)]. Moreover, the significance ofthe ratio between signals corresponding to lipids of type Iand II clearly indicates a change in their composition.Lipids of type I and II belong to different lipidicfragments, which include at least two types of moieties.These types of lipid fragments can be clearly identifiedin the 2D 1H,1H-TOCSY spectra and suggest thatthe distribution of lipid fragments change in the differentliver states. Interestingly, lysine seems to be anothercritical metabolite for discrimination between the
Figure 5. Aliphatic proton region of 1H water presaturation1D spectra of human liver tissue at 273K and spinning at4000Hz for patients with (A) non-pathological liver (sample16, 9mg), (B) hepatitis (sample 9, 2.9mg) and (C) cirrhosis(sample 7, 7.5mg). The spectra have not been normalized todifferent sample weights intentionally to include completelythe signals of free lipids and fatty acids
HR-MAS OF HUMAN LIVER NEEDLE BIOPSIES 9
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
different situations, although it is not visible in all spectra.In fact, the intensity ratio for signals of alanine and lysineshows strong differences among cirrhotic, hepatitisC-affected and non-pathological tissue [Fig. 6(C)]. Inaddition, some differences can also be detected for theratio between intensities for signals of monomers or shortpolymers of �- and �-glucose, although its significance isnot particularly good. Obviously, the significance of ourstatistical analysis must be considered in the particularcontext of this preliminary study, where the number ofcases presented is still small.
Overall, these trends may establish the basis fordeveloping diagnostic applications based on HR-MASspectroscopic data on human liver tissue. Althoughmetabolite quantification has not been performed in thework presented here, intensity ratios have been shown tobe a potentially useful tool for precise diagnosis of anumber of diseases. In addition, combined with otherdiagnosis tools, such as in vivo brain 1H MRS, thisapproach may provide information for the prediction ofpossible hepatic encephalopathy (30). In our experience,a more accurate study of the type of data presented herewould include a principal component analysis (PCA) ofthe 1D spectra, probably performed by the intensitybuckets approach. However, this type of study requiresa significantly larger number of cases and it is notappropriate for the data presented here. Future studiesbased on this work will provide robust and potent analysistools for overcoming these problems and provide usefuldiagnostic tools for liver disease using HR-MAS data.
CONCLUSION
The application of HR-MAS 1H NMR spectroscopy tothe study of intact human liver tissue obtained by needlebiopsy opens up new possibilities for the diagnosis of
liver diseases by biochemical profile determination. Theacquisition of well-resolved 1D 1H spectra on smallquantities of liver tissue was possible in less than15 min. Tissue degradation can be neglected during thefirst 4 h of data collection time by maintaining the sampleat temperatures as low as possible. The resonance assign-ments reported here allowed us to identify and quantifymany endogenous metabolites. The biochemical profileof liver metabolites obtained by HR-MAS NMR willincrease our knowledge of liver metabolism. Comparisonof NMR metabolic patterns with different controls, e.g.liver cells lines, may provide new discrimination tools fordifferent liver diseases. Although the observation ofglycogen signals offers new possibilities in the study ofglycogen metabolism, precise quantification requiresadditional investigation. A preliminary statistical analysison intensity ratios for many TOCSY signals providesthe basis to discriminate among cirrhotic, hepatitis C-affected and non-pathological liver tissue. Among others,these intensity ratio analyses show changes in lipidcomposition between cirrhotic and non-cirrhotic tissue.The technique may also be applied to other tissues, suchas prostate and breast, where needle biopsies are routi-nely performed in clinical environments.
Acknowledgements
Q3B. Martınez-Granados gratefully acknowledges theSpanish network ‘Imagen Medica Molecular y Multi-modalidad. Analisis y Tratamiento de Imagen Medica’,IM3 (Instituto de Salud Carlos III, G03/185) for apredoctoral fellowship. Thanks are due to the SCSIE ofthe University of Valencia for providing access to theNMR facility. We also thank Bruker Espana SA andBruker Biospin France for financial support as part of acollaboration with the University of Valencia. M. Piotto
Figure 6. Box diagram showing median (thick line), first quartile (gray box) andthe extreme values (thin lines) for TOCSY signal intensity ratios (A) �-glucose/lipidsI, (B) lipids II/lipids I and (C) alanine/lysine
Q3
10 B. MARTINEZ-GRANADOS ET AL.
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
and O. Assemat of Bruker Biospin France applicationslaboratory are gratefully acknowledged for useful discus-sions and technical support. The Association for theDevelopment of Research in NMR (ADIRM), the Ex-cellence Program of the Generalitat Valenciana (GVGRUPOS03/072) and the European Integrated ProjecteTUMOUR (FP6-2002-LIFESCIHEALTH 503094) areacknowledged for financial support. Finally, the financialsupport of the Red de Grupo de Encefalopatıa (03/155)del Instituto de Salud Carlos III, Ministerio de Educa-cion, Cultura y Deporte, Spain, is acknowledged.
REFERENCES
1. Cheng LL, Ma MJ, Becerra L, Ptak T, Tracey I, Lackner A,Gonzalez RG. Quantitative neuropathology by high resolutionmagic angle spinning proton magnetic resonance spectroscopy.Proc. Natl. Acad. Sci. USA 1997; 94: 6408–6413.
2. Moka D, Vorreuther R, Schicha H, Spraul M, Humpfer E,Lipinski M, Foxall PJD, Nicholson JK, Lindon JC. Magic anglespinning proton nuclear magnetic resonance spectroscopicanalysis of intact kidney tissue samples. Anal. Commun. 1997;34: 107–109.
3. Cheng LL, Chang IW, Smith BL, Gonzalez RG. Evaluatinghuman breast ductal carcinomas with high-resolution magic-anglespinning proton magnetic resonance spectroscopy. J. Magn.Reson. 1998; 135: 194–202.
4. Griffin JL, Williams HJ, Sang E, Nicholson JK. Abnormal lipidprofile of dystrophic cardiac tissue as demonstrated by one- andtwo-dimensional magic-angle spinning (1)H NMR spectroscopy.Magn. Reson. Med. 2001; 46: 249–255.
5. Cheng LL, Wu C, Smith MR, Gonzalez RG. Non-destructivequantitation of spermine in human prostate tissue samples usingHRMAS 1H NMR spectroscopy at 9.4 T. FEBS Lett. 2001; 494:112–116.
6. Millis KK, Maas WE, Cory DG, Singer S. Gradient, high-resolution, magic-angle spinning nuclear magnetic resonancespectroscopy of human adipocyte tissue. Magn. Reson. Med.1997; 38: 399–403.
7. Martınez-Bisbal MC, Martı-Bonmatı L, Piquer J, Revert A, FerrerP, Llacer JL, Piotto M, Assemat O, Celda B. 1H and 13C HR-MASspectroscopy of intact biopsy samples ex-vivo and in-vivo 1HMRS study of human high grade gliomas. NMR Biomed. 2004; 17:191–205.
8. Cheng LL, Lean CL, Bogdanova A, Wright SC Jr, Ackerman JL,Brady TJ, Garrido L. Enhanced resolution of proton NMR spectraof malignant lymph nodes using magic-angle spinning. Magn.Reson. Med. 1996; 36: 653–658.
9. Moka D, Vorreuther R, Schicha H, Spraul M, Humpfer E, LipinskiM, Foxall PJ, Nicholson JK, Lindon JC. Biochemical classifica-tion of kidney carcinoma biopsy samples using magic-angle-spinning 1H nuclear magnetic resonance spectroscopy. J. Pharm.Biomed. Anal. 1998; 17: 125–132.
10. Chen JH, Enloe BM, Fletcher CD, Cory DG, Singer S. Biochem-ical analysis using high-resolution magic angle spinning NMRspectroscopy distinguishes lipoma-like well-differentiated lipo-sarcoma from normal fat. J. Am. Chem. Soc. 2001; 123: 9200–9201.
11. Andrew ER, Eades RG. Removal of dipolar broadening of NMRspectra of solids by specimen rotation. Nature 1959; 183: 1802.
12. Cozzone PJ, Confort-Gouny S, Vion-DuryQ4. Tissue and cellextracts MRS. In Encyclopedia of NMR, Grant DM, Harris RK(eds). Wiley: New York, 1996; 4771–4775.
13. Garroway AN. Magic-angle sample spinning of liquids. J. Magn.Reson. 1982; 49: 168–171.
14. Weybright P, Millis K, Campbell N, Cory DG, Singer S. Gradient,high-resolution, magic angle spinning 1H nuclear magnetic re-sonance spectroscopy of intact cells. Magn. Reson. Med. 1998; 39:337–345.
15. Griffin JL, Bollard M, Nicholson JK, Bhakoo K. Spectral profilesof cultured neuronal and glial cells derived from HRMAS (1)HNMR spectroscopy. NMR Biomed. 2002; 15: 375–384.
16. Cheng LL, Chang IW, Louis DN, Gonzalez RG. Correlation ofhigh-resolution magic angle spinning proton magnetic resonancespectroscopy with histopathology of intact human brain tumorspecimens. Cancer Res. 1998; 58: 1825–1832.
17. Lehtimaki KK, Valonen PK, Griffin JL, Vaisanen TH, Grohn OH,Kettunen MI, Vepsalainen J, Yla-Herttuala S, Nicholson J, Kaup-pinen RA. Metabolite changes in BT4C rat gliomas undergoingganciclovir–thymidine kinase gene therapy-induced programmedcell death as studied by 1H NMR spectroscopy in vivo, ex vivoand in vitro. J. Biol. Chem. 2003; 278: 45915–45923.
18. Maxwell RJ, Martinez-Perez I, Cerdan S, Cabanas ME, Arus C,Moreno A, Capdevila A, Ferrer E, Bartomeus F, Aparicio A,Conesa G, Roda JM, Carceller F, Pascual JM, Howells SL,Mazucco R, Griffiths JR. Pattern recognition analysis of 1HNMR spectra from perchloric acid extracts of human brain tumorbiopsies. Magn. Reson. Med. 1998; 39: 869–877.
19. Waters NJ, Garrod S, Farrant RD, Haselden JN, Connor SC,Connelly J, Lindon JC, Holmes E, Nicholson JK. High-resolutionmagic angle spinning (1)H NMR spectroscopy of intact liver andkidney: optimization of sample preparation procedures and bio-chemical stability of tissue during spectral acquisition. Anal.Biochem. 2000; 282: 16–23.
20. Bollard ME, Garrod S, Holmes E, Lindon JC, Humpfer E, SpraulM, Nicholson JK. High-resolution (1)H and (1)H–(13)C magicangle spinning NMR spectroscopy of rat liver. Magn. Reson. Med.2000; 44: 201–207.
21. Hu JZ, Rommereim DN, Wind RA. High-resolution 1H NMRspectroscopy in rat liver using magic angle turning at a 1 Hzspinning rate. Magn. Reson. Med. 2002; 47: 829–836.
22. Piotto M, Elbayed K, Wieruszeski JM, Lippens G. Practicalaspects of shimming a high resolution magic angle spinningprobeQ5. J. Magn. Reson. 2005; 173: 84–89.
23. Q6Piotto M, Elbayed K, Wieruszeski JM, Lippens G. Practicalaspects of shimming a high resolution magic angle spinningprobe. J. Magn. Reson. 2005; 173: 84–89.
24. Goddard TD, Kneller DG. SPARKY 3. http://www.cgl.ucsf.edu/home/sparky/manual/, University of California, SanFrancisco (1994).
25. Seery JP, Taylor-Robinson SD. The application of magneticresonance spectroscopy to the study of hepatic encephalopathy.J. Hepatol. 1996; 25: 988–998.
26. Naegele T, Grodd W, Viebahn R, Seerger U, Klose U, Seitz D,Kaiser S, Mader I, Mayer J, Lauchart W, Gregor M, Voigt K. MRImaging and 1H spectroscopy of brain metabolites in hepaticencephalopathy: time-course of renormalization after liver trans-plantation. Radiology 2000; 216: 683–691.
27. Danielsen ER, Ross B. Magnetic Resonance Spectroscopy Diag-nosis of Neurological Diseases. Marcel Dekker: New York, 1999;32–33.
28. Beckwith-Hall BM, Nicholson JK, Nicholls AW, Foxall PJ,Lindon JC, Connor SC, Abdi M, Connelly J, Holmes E. Nuclearmagnetic resonance spectroscopic and principal components ana-lysis investigations into biochemical effects of three modelhepatotoxins. Chem. Res. Toxicol. 1998; 11: 260–272.
29. Wang Y, Bollard ME, Keun H, Antti H, Beckonert O, Ebbels TM,Lindon JC, Holmes E, Tang H, Nicholson JK. Spectral editing andpattern recognition methods applied to high-resolution magic-angle spinning 1H nuclear magnetic resonance spectroscopy ofliver tissues. Anal. Biochem. 2003; 323: 26–32.
30. Martinez-Granados B, Monleon D, Martinez-Bisbal MC, RodrigoJM, del Olmo J, Lluch P, Ferrandez A, Martı-Bonmatı L, Celda B.Quantitative analysis of HR-MAS spectra of human liver needlebiopsies and in vivo 1H MRS of brain and liver. Med. Biol. Eng.Comput. 2005; in pressQ7.
Q5
Q6
Q4
Q7
HR-MAS OF HUMAN LIVER NEEDLE BIOPSIES 11
Copyright # 2005 John Wiley & Sons, Ltd. NMR Biomed. 2005;18:1–11
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
Author Query Form (NBM/1005)
Special Instructions: Author please write responses to queries directly on Galley proofs and then fax
back. Alternatively please list responses in an e-mail.
Q1: Author: Weighing/weighed—correct?
Q2: Author: What is DSS?
Q3: Author: Please check these acknowledge-ments and the contract/grant sponsors on thefirst page—they are not exactly the same.Note that Fellowships etc. are not included infinancial support.
Q4: Author: Please give initials for the lastauthor.
Q5: Author: Is the beginning of the article titlenow correct?
Q6: Author: This is identical with Ref. 22.Either replace it with an alternative refer-ence, or delete and renumber subsequentreferences.
Q7: Author: Update?