Analysis of the fragile-X chromosome: localization and detection of the fragile site in high resolution preparations

Summary Fragile X chromosome preparations were analyzed at levels of up to 850 bands per haploid set. We were able to consistently sublocalize the Xqter fragile site to band Xq27.3 using high-resolution methods. Chromosome length versus the frequency of fragile X expression was also compared. The fragile site appeared at a higher percentage in more condensed chromosome preparations. The importance of this finding is that high-resolution chromosome preparations do not optimize fragile-X detection.     

Bacteriochlorophyll a-types in chromatophore and subchromatophore preparations from Rhodopseudomonas sphaeroides.

Comparison of absorption and circular dichroism (CD) spectra in the near infrared region was made with chromatophore and subchromatophore preparations obtained from Rhodopseudomonas sphaeroides. The 850 nm absorption band had a positive correlation with the 850 nm and 870 nm CD bands. The 800 nm and 870 nm absorption bands seemed not to correlate with any CD bands. Lipid contents in chromatophores and subchromatophores were measured. Lipids in membranes seemed to contribute to the appearance of the 870 nm absorption band, but not to that of the 800 nm and 850 nm absorption bands. The time courses of absorbance changes were compared at 800, 850, and 870 nm in detergent-treated chromatophores. Relative changes of absorbances differed from one another. The present results suggest that the three absorption bands are due to three different bacteriochlorophyll a-types and the 850 nm absorption band originates from exciton-coupling of bacteriochlorophyll a.     

Alkaline denaturation of the light-harvesting complex II from the purple bacterium Ectothiorhodospira sp.: kinetic evidence of the existence of the 780 nm upper exciton component of the B850 bacteriochlorophylls

The light-harvesting complex II of the purple bacteria has two strong near-infrared electronic absorption bands around 800 (B800) and 850 (B850) nm, arising from the Q(y)() transitions of the bacteriochlorophyll a. In the present work, high concentrations of NaOH were used to study the destabilization of the complex of the Ectothiorhodospira sp. The majority of the bacteriochlorophylls were monomerized within 90 min of treatment. However, the kinetic patterns of the two near-infrared bands were remarkably different. After an instantaneous blue shift from 853 to 828 nm, B850 showed a first-order monomerization with a rate constant of -0.016 min(-1). This instantaneous blue shift was previously attributed to the deprotonation of a lysine and was independent of the monomerization process. The observed native B800 is in fact composed of two bands, one at 796 nm and the other at 780 nm. The band absorbing at 780 nm red shifted also instantaneously to 786-788 nm and then disappeared in a first-order process as B850. The other band absorbing at 796 nm has a two-step process of monomerization; after a rapid conversion a slower first-order process occurred with a rate constant of -0.025 min(-1). The similarity between the kinetic behaviors of B850 and the 780 nm band indicated a strong relationship between these two bands. Our interpretation of the results considers the 780 nm band as the upper exciton component of the B850 bacteriochlorophylls.     

Characteristics and variations of B-type DNA conformations in solution: a quantitative analysis of Raman band intensities of eight DNAs

Raman spectra were obtained from four bacterial DNAs varying in GC content and four periodic DNA polymers in 0.1 M NaCl at 25 degrees C. A curve fitting procedure was employed to quantify and compare Raman band characteristics (peak location, height, and width) from 400 to 1600 cm-1. This procedure enabled us to determine the minimum number of Raman bands in regions with overlapping peaks. Quantitative comparison of the Raman bands of the eight DNAs provided several new results. All of the DNAs examined required bands near 809 (+/- 7) and 835 (+/- 5) cm-1 to accurately reproduce the experimental spectra. Since bands at these frequencies are associated with A-family and B-family conformations, respectively, this result indicates that all DNAs in solution have a mixture of conformations on the time scale of the Raman scattering process. Band characteristics in the 800-850-cm-1 region exhibited some dependence on CG content and base pair sequence. As previously noted by Thomas and Peticolas [Thomas, G. A., & Peticolas, W. L. (1983) J. Am. Chem. Soc. 105, 993], the poly[d(A)].poly[d(T)] spectra were qualitatively distinct in this region. The A-family band is clearly observed at 816 cm-1. The intensity of this band and that of the B-family band at 841 cm-1 were similar, however, to intensities in the natural DNA spectra. Three bands at 811, 823, and 841 cm-1 were required to reproduce the 800-850-cm-1 region of the poly[d(A-T)].poly[d(A-T)] spectra. This may indicate the presence of three backbone conformations in this DNA polymer. Analysis of intensity vs. GC content for 42 Raman bands confirmed previous assignments of base and backbone vibrations and provided additional information on a number of bands.     

BrdU处理的鱼类染色体高分辨G-带带型分析

本文应用鱼类染色体高分辨G-带技术,重点将黄鳝培养细胞具不同长度染色体的正中期分裂相做成G-带核型加以比较分析。随着染色体长度的增加,带纹数目也增加。但增加是有限度的。染色体带纹数目的增加,明显地表现在深染带再分为若干亚带。当染色体从前期向中、后期过渡收缩变短时,一些亚带融合为原来数目的带。染色体上各个带的收缩程度、收缩时间是不均等的。实验证明大剂量的BrdU不仅能阻断鱼类细胞于中S期,也可使染色体伸长、小剂量的伸长作用不明显。最后讨论了BrdU处理与G-显带的关系、染色体带纹数目相对恒定以及染色体伸长缩短问题。     

Analysis of the pigment content of an antenna pigment-protein complex from three strains of Rhodopseudomonas sphaeroides.

The pigment content of a B800-850 light-harvesting pigment-protein complex isolated from three different stains of Rhodopseudomonas sphaeroides has been determined. In each case the ratio of carotenoid to bacteriochlorophyll present is very nearly 1 : 3 an no specificity with regard to carotenoid type was observed. The fourth derivative of the infra-red absorption bands of the complex was determined and it is concluded that the minimal functional unit of B800-850 complex consists of 1 carotenoid molecule and three bacteriochlorophyll molecules. The data presented here, together with the previous study of Austin, (Austin, L.A. (1976) Ph.D. Thesis, University of California at Berkeley, Lawrence Berkeley Laboratory Report No. LBL 5512) suggest that the 800 nm absorption band represents one of these bacteriochlorophyll molecules while the remaining two bacteriochlorophylls are responsible for the 850 nm band. The absorption spectra and circular dichroism spectra of the complexes suggests that their structure has not been greatly altered during the purification.     

The hierarchically organized splitting of chromosome bands into sub-bands analyzed by multicolor banding (MCB)

To clarify the nature of chromosome sub-bands in more detail, the multicolor banding (MCB) probe-set for chromosome 5 was hybridized to normal metaphase spreads of GTG band levels at approximately 850, approximately 550, approximately 400 and approximately 300. It could be observed that as the chromosomes became shorter, more of the initial 39 MCB pseudo-colors disappeared, ending with 18 MCB pseudo-colored bands at the approximately 300-band level. The hierarchically organized splitting of bands into sub-bands was analyzed by comparing the disappearance or appearance of pseudo-color bands of the four different band levels. The regions to split first are telomere-near, centromere-near and in 5q23-->q31, followed by 5p15, 5p14, and all GTG dark bands in 5q apart from 5q12 and 5q32 and finalized by sub-band building in 5p15.2, 5q21.2-->q21.3, 5q23.1 and 5q34. The direction of band splitting towards the centromere or the telomere could be assigned to each band separately. Pseudo-colors assigned to GTG-light bands were resistant to band splitting. These observations are in concordance with the recently proposed concept of chromosome region-specific protein swelling.     

Effects of acid pH and urea on the spectral properties of the LHII antenna complex from the photosynthetic bacterium Ectothiorhodospira sp.

The aim of this study was to investigate the spectral modifications of the LHII antenna complex from the purple bacterium Ectothiorhodospira sp. upon acid pH titration both in the presence and absence of urea. A blue shift specifically and reversibly affected the B850 band around pH 5.5-6.0 suggesting that a histidine residue most probably participated in the in vivo absorption red shifting mechanism. This transition was observed in the presence and absence of urea. Under strong chaotropic conditions, a second transition occurred around pH 2.0, affecting the B800 band irreversibly and the B850 reversibly. Under these conditions a blue shift from 856 to 842 nm occurred and a new and strong circular dichroism signal from the new 842 nm band was observed. Reverting to the original experimental conditions induced a red shift of the B850 band up to 856 nm but the circular dichroism signal remained mostly unaffected. Under the same experimental conditions, i.e. pH 2.1 in the presence of urea, part of the B800 band was irreversibly destroyed with concomitant appearance of a band around 770 nm due to monomeric bacteriochlorophyll from the disrupted B800. Furthermore, Gaussian deconvolution and second derivative of the reverted spectra at pH 8.0 after strong-acid treatment indicated that the new B850 band was actually composed of two bands centered at 843 and 858 nm. We ascribed the 858 nm band to bacteriochlorophylls that underwent reversible spectral shift and the 843 nm band to oligomeric bacteriopheophytin formed from a part of the B850 bacteriochlorophyll. This new oligomer would be responsible for the observed strong and mostly conservative circular dichroism signal. The presence of bacteriopheophytin in the reverted samples was definitively demonstrated by HPLC pigment analysis. The pheophytinization process progressed as the pH decreased below 2.1, and at a certain point (i.e. pH 1.5) all bacteriochlorophylls, including those from the B800 band, became converted to oligomeric bacteriopheophytin, as shown by the presence of a single absorption band around 843 nm and by the appearance of a single main elution peak in the HPLC chromatogram which corresponded to bacteriopheophytin.     

Low-intensity pump-probe measurements on the B800 band of Rhodospirillum molischianum

We have measured low-intensity, polarized one-color pump-probe traces in the B800 band of the light-harvesting complex LH2 of Rhodospirillum molischianum at 77 K. The excitation/detection wavelength was tuned through the B800 band. A single-wavelength and a global target analysis of the data were performed with a model that accounts for excitation energy transfer among the B800 molecules and from B800 to B850. By including the anisotropy of the signals into the fitting procedure, both transfer processes could be separated. It was estimated in the global target analysis that the intra-B800 energy transfer, i.e., the hopping of the excitation from one B800 to another B800 molecule, takes approximately 0.5 ps at 77 K. This transfer time increases with the excitation/detection wavelength from 0.3 ps on the blue side of the B800 band to approximately 0.8 ps on the red side. The residual B800 anisotropy shows a wavelength dependence as expected for energy transfer within an inhomogeneously broadened cluster of weakly coupled pigments. In the global target analysis, the transfer time from B800 to B850 was determined to be approximately 1.7 ps at 77 K. In the single-wavelength analysis, a speeding-up of the B800 --> B850 energy transfer rate toward the blue edge of the B800 band was found. This nicely correlates with the proposed position of the suggested high-exciton component of the B850 band acting as an additional decay channel for B800 excitations.     

Assembly of LH2 light-harvesting complexes in Rhodopseudomonas palustris cells illuminated by blue and red light

We investigated the formation of the B800-850 complex in cells of the bacterium Rhodopseudomonas palustris AB illuminated by red and blue light under anaerobic growth conditions. Under red illumination, the B800-850 complex was assembled with a reduced absorption band at 850 nm. The results of re-electrophoresis of the B800-850 complex and oxidation in the presence of potassium iridate suggest its heterogeneity. It may be a mixture of two complexes (B800 and B800-850). The B800-850 complex lacks the capacity for conformational transitions if assembled under blue illumination. Accordingly, the light-harvesting complex assembled in the blue light contains polypeptides that are not synthesized under normal conditions or at increased or decreased light intensities. The mechanism of regulation of the synthesis of the polypeptides of light-harvesting B800-850 complex and its dependence on the spectral composition of the light is discussed.     

High-resolution analysis of DNA replication domain organization across an R/G-band boundary.

Establishing how mammalian chromosome replication is regulated and how groups of replication origins are organized into replication bands will significantly increase our understanding of chromosome organization. Replication time bands in mammalian chromosomes show overall congruency with structural R- and G-banding patterns as revealed by different chromosome banding techniques. Thus, chromosome bands reflect variations in the longitudinal structure and function of the chromosome, but little is known about the structural basis of the metaphase chromosome banding pattern. At the microscopic level, both structural R and G bands and replication bands occupy discrete domains along chromosomes, suggesting separation by distinct boundaries. The purpose of this study was to determine replication timing differences encompassing a boundary between differentially replicating chromosomal bands. Using competitive PCR on replicated DNA from flow-sorted cell cycle fractions, we have analyzed the replication timing of markers spanning roughly 5 Mb of human chromosome 13q14.3/q21.1. This is only the second report of high-resolution analysis of replication timing differences across an R/G-band boundary. In contrast to previous work, however, we find that band boundaries are defined by a gradient in replication timing rather than by a sharp boundary separating R and G bands into functionally distinct chromatin compartments. These findings indicate that topographical band boundaries are not defined by specific sequences or structures.     

Assembly of LH2 light-harvesting complexes in Rhodopseudomonas palustris cells illuminated by blue and red light

We investigated the formation of the B800-850 complex in cells of the bacterium Rhodopseudomonas palustris AB illuminated by red and blue light under anaerobic growth conditions. Under red illumination, the B800-850 complex was assembled with a reduced absorption band at 850 nm. The results of re-electrophoresis of the B800-850 complex and oxidation in the presence of potassium iridate suggest its heterogeneity. It may be a mixture of two complexes (B800 and B800-850). The B800-850 complex lacks the capacity for conformational transitions if assembled under blue illumination. Accordingly, the light-harvesting complex assembled in the blue light contains polypeptides that are not synthesized under normal conditions or at increased or decreased light intensities. The mechanism of regulation of the synthesis of the polypeptides of light-harvesting the B800-850 complex and its dependence on the spectral composition of the light is discussed.     

Localization of the human salivary protein complex (SPC) to chromosome band 12p13.2

In situ hybridization of a 3H-labeled probe containing a fragment from PRP-1, a genomic clone with human salivary proline-rich protein gene sequences, revealed significant labeling on the short arm of human chromosome 12 in metaphase preparations from two individuals. Fifty-three percent of metaphases exhibited labeling on one or both chromosomes 12. Additional cells scored at the 850-1,000 band level revealed a significant proportion (52% [32/61] grains, p less than 0.005) of the labeled sites on chromosome 12 to be on band 12p13.2. This probe for a human salivary proline-rich protein gene fragment, probably PMS, is from a cluster of 13 linked genes designated as the human salivary protein complex (SPC). Studies of the DNA of human-mouse somatic-cell hybrids have assigned the SPC to chromosome 12, but have not provided a regional localization (Azen et al, 1985). This paper reports the localization of the SPC to a specific chromosomal band, 12p13.2.     

Identification of the carotenoid present in the B-800–850 antenna complex from Rhodopseudomonas capsulata as that which responds electrochromically to transmembrane electric fields

Mild proteolysis of Rhodopseudomonas capsulata chromatophores results in a parallel loss of the 800 nm bacteriochlorophyll absorption band and a blue shift in the carotenoid absorption bands associated with the B-800–850 light-harvesting complex. Both the light-induced and the salt-induced electrochromic carotenoid band shift disappear in parallel to the loss of the 800 nm bacteriochlorophyll absorption upon pronase treatment of chromatophores. During the time required for the loss of the 800 nm bacteriochlorophyll absorption and the loss of the electrochromic carotenoid band shift photochemistry is not inhibited and the ionic conductance of the membrane remains very low. We conclude that the carotenoid associated with the B-800–850 light-harvesting complex is the one that responds electrochromically to the transmembrane electric field. Analysis of the pigment content of Rps. capsulata chromatophores indicates that all of the carotenoid may be accounted for in the well defined pigment-protein complexes.     

High-resolution idiogram of Giemsa R-banded human prophase chromosomes

The schematic representation of RHG-banded chromosomes (R-banding was produced by heat denaturation followed by Giemsa staining (RHG) in the 850-band range per haploid set, was prepared showing the relative position, the specific size, and the characteristic staining intensity for each band. To this idiogram was adapted the new International Standard Cytogenetic Nomenclature. Our aim was to produce a realistic idiogram which could help in the preparation of R-banded prophase karyotypes and in the localization of chromosomal rearrangements. A comparative analysis of bands at prophase and metaphase revealed certain aspects of the dynamics involved in chromosome condensation and in R-band organization. The effect of chromosome elongation on the appearance of R-bands within heterochromatic regions has also been discussed.     

B800-->B850 energy transfer mechanism in bacterial LH2 complexes investigated by B800 pigment exchange

Femtosecond transient absorption measurements were performed on native and a series of reconstituted LH2 complexes from Rhodopseudomonas acidophila 10050 at room temperature. The reconstituted complexes contain chemically modified tetrapyrrole pigments in place of the native bacteriochlorophyll a-B800 molecules. The spectral characteristics of the modified pigments vary significantly, such that within the B800 binding sites the B800 Q(y) absorption maximum can be shifted incrementally from 800 to 670 nm. As the spectral overlap between the B800 and B850 Q(y) bands decreases, the rate of energy transfer (as determined by the time-dependent bleaching of the B850 absorption band) also decreases; the measured time constants range from 0.9 ps (bacteriochlorophyll a in the B800 sites, Q(y) absorption maximum at 800 nm) to 8.3 ps (chlorophyll a in the B800 sites, Q(y) absorption maximum at 670 nm). This correlation between energy transfer rate and spectral blue-shift of the B800 absorption band is in qualitative agreement with the trend predicted from F?rster spectral overlap calculations, although the experimentally determined rates are approximately 5 times faster than those predicted by simulations. This discrepancy is attributed to an underestimation of the electronic coupling between the B800 and B850 molecules.     

Identification and designation of telocentric chromosomes in barley by means of Giemsa N-banding technique

Summary Seven complete chromosomes and nine telocentric chromosomes in telotrisomics of barley (Hordeum vulgare L.) were identified and designated by an improved Giemsa N-banding technique. Karyotype analysis and Giemsa N-banding patterns of complete and telocentric chromosomes at somatic late prophase, prometaphase and metaphase have shown the following results: Chromosome 1 is a median chromosome with a long arm (Telo 1L) carrying a centromeric band, while short arm (Telo 1S) has a centromeric band and two intercalary bands. Chromosome 2 is the longest in the barley chromosome complement. Both arms show a centromeric band, an intercalary band and two faint dots on each chromatid at middle to distal regions. The banding pattern of Telo 2L (a centromeric and an intercalary band) and Telo 2S (a centromeric, two intercalary and a terminal band) corresponded to the banding pattern of the long and short arm of chromosome 2. Chromosome 3 is a submedian chromosome and its long arm is the second longest in the barley chromosome complement. Telo 3L has a centromeric (fainter than Telo 3S) and an intercalary band. It also shows a faint dot on each chromatid at distal region. Telo 3S shows a dark centromeric band only. Chromosome 4 is the most heavily banded one in barley chromosome complement. Both arms showed a dark centromeric band. Three dark intercalary bands and faint telomeric dot were observed in the long arm (4L), while two dark intercalary bands in the short arm (4S) were arranged very close to each other and appeared as a single large band in metaphase chromosomes. A faint dot was observed in each chromatid at the distal region in the 4S. Chromosome 5 is the smallest chromosome, which carries a centromeric band and an intercalary band on the long arm. Telo 5L, with a faint centromeric band and an intercalary band, is similar to the long arm. Chromosomes 6 and 7 are satellited chromosomes showing mainly centromeric bands. Telo 6S is identical to the short arm of chromosome 6 with a centromeric band. Telo 3L and Telo 4L were previously designated as Telo 3S and Telo 4S based on the genetic/linkage analysis. However, from the Giemsa banding pattern it is evident that these telocentric chromosomes are not correctly identified and the linkage map for chromosome 3 and 4 should be reversed. One out of ten triple 2S plants studied showed about 50% deficiency in the distal portion of the short arm. Telo 4L also showed a deletion of the distal euchromatic region of the long arm. This deletion (32%) may complicate genetic analysis, as genes located on the deficient segment would show a disomic ratio. It has been clearly demonstrated that the telocentric chromosomes of barley carry half of the centromere. Banding pattern polymorphism was attributed, at least partly, to the mitotic stages and differences in techniques.Contribution from the Department of Agronomy and published with the approval of the Director of the Colorado State University Experiment Station as Scientific Series Paper No. 2730. This research was supported in part by the USDA/SEA Competitive Research Grant 5901-0410-9-0334-0, USDA/ SEA-CSU Cooperative Research Grant 12-14-5001-265 and Colorado State University Hatch Project. This paper was presented partly at the Fourth International Barley Genetics Symposium, Edinburgh, Scotland, July 22–29, 1981     

Analysis of the pigment content of an antenna pigment-protein complex from three strains of Rhodopseudomonas sphaeroides

The pigment content of a B800–850 light-harvesting pigment-protein complex isolated from three different stains of Rhodopseudomonas sphaeroides has been determined. In each case the ratio of carotenoid to bacteriochlorophyll present is very nearly 1 : 3 an no specificity with regard to carotenoid type was observed.The fourth derivative of the infra-red absorption bands of the complex was determined and it is concluded that the minimal functional unit of B800–850 complex consists of 1 carotenoid molecule and three bacteriochlorophyll molecules. The data presented here, together with the previous study of Austin, (Austin, L.A. (1976) Ph.D. Thesis, University of California at Berkeley, Lawrence Berkeley Laboratory Report No. LBL 5512) suggest that the 800 nm absorption band represents one of these bacteriochlorophyll molecules while the remaining two bacteriochlorophylls are responsible for the 850 nm band.The absorption spectra and circular dichroism spectra of the complexes suggests that their structure has not been greatly altered during the purification.     

High pressure studies of energy transfer and strongly coupled bacteriochlorophyll dimers in photosynthetic protein complexes

High pressure is used with hole burning and absorption spectroscopies at low temperatures to study the pressure dependence of the B800B850 energy transfer rate in the LH2 complex of Rhodobacter sphaeroides and to assess the extent to which pressure can be used to identify and characterize states associated with strongly coupled chlorophyll molecules. Pressure tuning of the B800–B850 gap from 750 cm^\s-1 at 0.1 MPa to 900 cm^-1 at 680 MPa has no measurable effect on the 2 ps energy transfer rate of the B800–850 complex at 4.2 K. An explanation for this resilience against pressure, which is supported by earlier hole burning studies, is provided. It is based on weak coupling nonadiabatic transfer theory and takes into account the inhomogeneous width of the B800–B850 energy gap, the large homogeneous width of the B850 band from exciton level structure and the Franck-Condon factors of acceptor protein phonons and intramolecular BChl a modes. The model yields reasonable agreement with the 4.2 K energy transfer rate and is consistent with its weak temperature dependence. It is assumed that it is the C₉-ring exciton levels which lie within the B850 band that are the key acceptor levels, meaning that BChl a modes are essential to the energy transfer process. These ring exciton levels derive from the strongly allowed lowest energy component of the basic B850 dimer. However, the analysis of B850s linear pressure shift suggests that another Förster pathway may also be important. It is one that involves the ring exciton levels derived from the weakly allowed upper component of the B850 dimer which we estimate to be quasi-degenerate with B800. In the second part of the paper, which is concerned with strong BChl monomer-monomer interactions of dimers, we report that the pressure shifts of B875 (LH2), the primary donor absorption bands of bacterial RC (P870 of Rb. sphaeroides and P960 of Rhodopseudomonas viridis) and B1015 (LH complex of Rps. viridis) are equal and large in value (-0.4 cm⁰¹/MPa at 4.2 K) relative to those of isolated monomers in polymers and proteins (< -0.1 cm⁰¹/MPa). The shift rate for B850 at 4.2 K is-0.28 cm^–1/MPa. A model is presented which appears to be capable of providing a unified explanation for the pressure shifts.Abbreviations B800 BChl antenna band absorbing (at room temperature) at 800 nm (B850, B875, B1015 are defined similarly) - CD circular dichroism - FC factor Franck-Condon factor - FMO comple Fenna-Matthews-Olson complex - L-S theory Laird-Skinner theory - LH1 core light-harvesting complex of the BChl antenna complexes - LH2 peripheral light-harvesting complex of the BChl antenna complexes - NPHB non-photochemical hole burning - P960 absorption band of special pair of Rhodopseudomonas viridis absorbing at 960 nm (room temperature). P870 of Rhodobacter sphaeroides is defined similarly - QM/MM results quantum mechanical/molecular mechanical results - RC reaction center - ZPH zero phonon hole     

Observation of the energy-level structure of the low-light adapted B800 LH4 complex by single-molecule spectroscopy

Low-light adapted B800 light-harvesting complex 4 (LH4) from Rhodopseudomonas palustris is a complex in which the arrangement of the bacteriochloropyll a pigments is very different from the well-known B800-850 LH2 complex. For bulk samples, the main spectroscopic feature in the near-infrared is the occurrence of a single absorption band at 802 nm. Single-molecule spectroscopy can resolve the narrow bands that are associated with the exciton states of the individual complexes. The low temperature (1.2 K) fluorescence excitation spectra of individual LH4 complexes are very heterogeneous and display unique features. It is shown that an exciton model can adequately reproduce the polarization behavior of the complex, the experimental distributions of the number of observed peaks per complex, and the widths of the absorption bands. The results indicate that the excited states are mainly localized on one or a few subunits of the complex and provide further evidence supporting the recently proposed structure model.