Effects of the deletion of the Escherichia coli frataxin homologue CyaY on the respiratory NADH:ubiquinone oxidoreductase
© Pohl et al. 2007
Received: 15 November 2006
Accepted: 24 July 2007
Published: 24 July 2007
Skip to main content
© Pohl et al. 2007
Received: 15 November 2006
Accepted: 24 July 2007
Published: 24 July 2007
Frataxin is discussed as involved in the biogenesis of iron-sulfur clusters. Recently it was discovered that a frataxin homologue is a structural component of the respiratory NADH:ubiquinone oxidoreductase (complex I) in Thermus thermophilus. It was not clear whether frataxin is in general a component of complex I from bacteria. The Escherichia coli homologue of frataxin is coined CyaY.
We report that complex I is completely assembled to a stable and active enzyme complex equipped with all known iron-sulfur clusters in a cyaY mutant of E. coli. However, the amount of complex I is reduced by one third compared to the parental strain. Western blot analysis and live cell imaging of CyaY engineered with a GFP demonstrated that CyaY is located in the cytoplasm and not attached to the membrane as to be expected if it were a component of complex I.
CyaY plays a non-essential role in the assembly of complex I in E. coli. It is not a structural component but may transiently interact with the complex.
The NADH:ubiquinone oxidoreductase, also known as respiratory complex I, is the entry point for electrons in the respiratory chains of most bacteria and many eucaryotes. It links the electron transfer from NADH to ubiquinone with the translocation of protons across the membrane. In doing so, complex I establishes a proton motive force required for energy consuming processes [1–5]. One FMN and, depending on the species, eight to nine iron-sulfur (Fe/S) clusters participate in the electron transfer reaction. Generally, the bacterial complex I consists of 14 different subunits called NuoA through NuoN (or Nqo1 through Nqo14; [5–10]). In a few bacteria such as Escherichia coli and Aquifex aeolicus the genes nuoC and nuoD are fused resulting in a complex consisting of 13 subunits. Seven (or six, see above) peripheral proteins including those that bear all known redox groups build up the so-called peripheral arm of the complex, which extends into the aqueous medium. The residual seven subunits are hydrophobic proteins and build the membrane arm of the complex. The arrangement of both arms of the complex have been visualized by means of electron microscopy [11, 12].
Recently, the structure of the peripheral arm of the complex from Thermus thermophilus was resolved at 3.3 Å resolution . This pathbreaking study revealed the unexpected presence of a 15th subunit in the T. thermophilus complex [13, 14]. The subunit was coined Nqo15 and exhibits structural similarity to the frataxin family. Nqo15 shows a 2.5 Å RMSD to the structure of CyaY, the E. coli frataxin homologue [13, 15]. Despite the similar three-dimensional fold, the sequence similarity of Nqo15 to members of the frataxin family is very low. Homologues of Nqo15 were only detected in close relatives of Thermus, such as Deinococcus species [13, 14].
Frataxin was first recognized in patients suffering from Friedreich's ataxia [16–18]. Its loss in patients results in a neurodegenerative disease due to an unbalanced iron homeostasis and oxidative damage [16–19]. The exact physiological function of frataxin is still under debate. Frataxin has been shown to specifically albeit weakly bind iron  and it was discussed that it is involved in the assembly of Fe/S clusters [21–23]. Recently, it was shown in E. coli that CyaY binds to IscS, the cysteine desulfurase of the ISC-system, and delivers iron to the scaffold protein IscU . The deletion of cyaY in E. coli had no effect on the cellular iron content and its sensitivity to oxidants  but it was shown in Salmonella enterica that the deletion of cyaY in combination with other specific lesions resulted in severe metabolic defects .
In this study, we used a cyaY deletion mutant to investigate whether or not CyaY is a structural component of the E. coli complex I. Our data show that this is not the case but that CyaY is most likely a non-essential component for the assembly of the E. coli complex I.
E. coli contains two membrane-bound NADH dehydrogenases, the energy-converting complex I and a non-energy-converting, alternative NADH dehydrogenase . While NADH is a substrate for both enzymes the artificial substrate deamino (d)-NADH is only a poor substrate for the alternative dehydrogenase and used to descriminate both enzymes [27, 28]. The NADH/ferricyanide oxidoreductase activity of the cytoplasmic membranes was virtually identical in the parental strain and the cyaY deletion strain (Table 1). Thus, the total amount of complex I and the alternative NADH dehydrogenase did not differ in the strains.
Catalytic activities of cytoplasmic membranes from the parental and the cyaY deletion strain. The data are the mean of three independent measurements.
NADH/ferricyanide oxidoreductase activity
NADH oxidase activity
Inhibition by annonin VI
d-NADH oxidase activity
Inhibition by annonin VI
[μmol· min-1 ·mg-1]
[μmol· min-1 ·mg-1]
[μmol· min-1· mg-1]
2.6 ± 0.5
0.29 ± 0.03
0.18 ± 0.01
2.7 ± 0.7
0.33 ± 0.06
0.13 ± 0.01
Succinate oxidase activity
Inhibition by malonate
[μmol· min-1 ·mg-1]
0.12 ± 0.01
0.09 ± 0.01
The physiological NADH oxidase activity of the parental strain was inhibited by 52% by annonin VI, which selectively blocks complex I (Table 1), as observed with other E. coli strains . The inhibitor-insensitive activity derived from the alternative NADH dehydrogenase. The d-NADH oxidase activity of this strain, which stems from complex I, was 62% of the activity with NADH as substrate and completely inhibited by annonin VI (Table 1).
Using the cyaY deletion strain, an NADH oxidase activity similar to the parental strain was measured, but the inhibition of the activity by annonin VI was approximately 20% (Table 1). This indicated a higher amount of the alternative NADH dehydrogenase in the strain. The d-NADH oxidase activity of the mutant membranes was 39% of the activity with NADH as substrate and 28% lower than the rate measured with the membranes from the parental strain (Table 1). As in the membranes from the parental strain, the d-NADH oxidase activity was fully sensitive to annonin VI, indicating that the d-NADH oxidase activity in this strain derived from complex I. The data demonstrate that a functionally active complex I is present in the cytoplasmic membranes of the cyaY deletion mutant, but in a lesser amount.
To investigate whether the cyaY deletion had an effect on other Fe/S cluster containing complexes of the E. coli respiratory chain, we measured the succinate oxidase activity of the cytoplasmic membranes from both strains. This activity was fully sensitive to malonate (Table 1), a specific inhibitor of the E. coli succinate dehydrogenase, the respiratory complex II. The succinate oxidase activity in the mutant membranes was 75% of the activity measured in the membranes of the parental strain (Table 1). Thus, the cyaY deletion mutant contains an active succinate dehydrogenase but in lesser amounts. From these measurements it is concluded that the amount of complex I containing nine Fe/S cluster is reduced by one third and the amount of complex II containing three Fe/S clusters by one quarter in the cyaY deletion strain.
No enhanced NADH/ferricyanide oxidoreductase activity was detected in fractions of the gradient corresponding to higher molecular masses, revealing that the complex from the cyaY deletion strain showed no tendency to aggregate (Fig. 1). No NADH/ferricyanide oxidoreductase activity was detected in fractions 7 to 10 corresponding to the position of a soluble fragment of the complex [30, 32], indicating that the complex from the deletion strain did not disintegrate. The NADH/ferricyanide oxidoreductase activity around fraction 5 is due to the alternative NADH dehydrogenase . The activity of this fraction was approximately doubled in the gradient obtained from the cyaY deletion strain, demonstrating an enhanced amount of the alternative NADH dehydrogenase in the mutant strain as already indicated by its NADH oxidase activity (Table 1). Thus, a stable complex I was properly assembled in the absence of CyaY.
Preparation of complex I. Isolation of E. coli complex I from strain BW25113 cyaY::nptI starting from 76 g cells (wet weight)*.
NADH/ferricyanide oxidoreductase activity
total [μmol· min-1]
specific [μmol· min-1 ·mg-1]
The CyaY protein decorated with a His-tag was overproduced and purified by affinity chromatography by means of His Spin-Trap on Ni-Sepharose. SDS-PAGE showed the presence of two proteins in the preparation with apparent molecular masses of 12 and 25 kDa, respectively. The molecular mass of CyaY as deduced from its DNA sequence is 12.2 kDa . Thus, the two proteins were attributed to the monomeric and the dimeric form of CyaY, due to a different load with iron . The electrophoretic mobility of the monomeric form of CyaY did not match the mobility of any of the complex I subunits of the preparation from the parental strain (Fig. 3). The CyaY homologue of T. thermophilus, Nqo15, was detected in its monomeric form after SDS-PAGE of the peripheral arm of the complex . Thus, none of the obtained data indicated that CyaY could be an integral component of the E. coli complex I.
The overproduced CyaY modified with a His-tag was capable of binding 6.7 ± 0.5 mol Fe3+/mol CyaY under aerobic conditions. It has been shown that recombinant CyaY can bind eight Fe3+ . Thus, failure to detect CyaY within the membrane fraction is not due to a distorted Fe-binding of the recombinant protein.
The previously unrecognized protein Nqo15 was found to be a structural component of the T. thermophilus complex I [13, 14]. This soluble protein was co-purified with the preparation of the peripheral arm of the respiratory complex I from the membrane fraction of the disrupted cells. It co-eluted with the peripheral arm from four different chromatographic steps, and SDS-PAGE of the preparation revealed that it was present in stoichiometric amounts . Finally, the crystal structure of the peripheral arm unambiguously showed that this subunit is an integral part of the T. thermophilus complex I .
The genes comprising the structural subunits of the bacterial complex I are for the most part organized in one operon or in one or more gene clusters [6, 7]. The complex I gene clusters in Rhodobacter capsulatus  and Paracoccus denitrificans  contain additional ORFs that are not related to each other. These ORFs do not encode known complex I subunits and it has been shown that a disruption of the ORFs has no influence on the assembly and the enzymatic activity of complex I [5, 39]. The gene coding the CyaY homologue in T. thermophilus is neither located within nor in close proximity to the complex I gene cluster. Thus, the finding of this protein as an integral complex I component was completely unexpected .
The assignment of Nqo15 to the family of frataxin single domain proteins (cd00503) was not obvious from its primary structure. Solely the crystal structure revealed that Nqo15 exhibits the typical fold of a member of the frataxin family . Electron density for bound iron was not detected in the structure, implying that the metal might have been lost during purification. Here, we have shown that the frataxin homologue of E. coli CyaY is not a structural component of complex I. The preparation of complex I from the wild type did not contain a subunit of an apparent molecular mass corresponding to the molecular mass of CyaY (Fig. 3). The complex I was fully assembled in the membrane of the mutant and contains all subunits and known cofactors (Table 1; Figs. 1, 3, and 4). Its physiological NADH:decyl-ubiquinone oxidoreductase activity was indistinguishable from the activity of the complex I preparation from wild type. We were able to verify the analyses in vivo, because GFP-labelled CyaY did not associate with the membrane like NuoJ, but showed a cytosolic localization. Thus, it is reasonable to assume that CyaY is not a structural component of the E. coli complex I.
An important finding in this study is that the amount of complex I in the mutant was reduced by one third as derived from the inhibitor-sensitive portion of the NADH oxidase activity and the d-NADH oxidase activity of the membranes (Table 1) as well as from the sucrose gradient centrifugation of detergent extracts (Fig. 1). The overall NADH dehydrogenase activity of the membranes did not change as indicated by the unaltered NADH/ferricyanide oxidoreductase activity (Table 1). These data demonstrate that the amount of the alternative NADH dehydrogenase is increased approximately two-fold. Thus, the ratio of complex I to the alternative enzyme changed due to the deletion of cyaY. This might be due to a change of the metabolic and/or redox state of the mutant cells. However, it is well known that mutations in the complex I genes leading to a decreased amount of the complex or the assembly of an inactive complex result in a two-fold enhanced production of the alternative enzyme [10, 27, 29, 30]. Therefore, it seems rather unlikely that the deletion of cyaY changed the properties of the cell leading to a subsequent change of the ratio of complex I to the alternative NADH dehydrogenase.
The deletion of cyaY led to a decrease of the complex I and complex II content of the cytoplasmic membrane by approximately one third and one quarter, respectively (Table 1). This implies a transient interaction between CyaY and the complexes of the E. coli respiratory chain. It is possible that CyaY is involved in either biogenesis or repair of the Fe/S clusters of the complexes containing Fe/S clusters. The stoichiometry of the individual Fe/S clusters in complex I were not distinguishable in the preparation from wild type and the cyaY deletion strain (Fig. 4). If CyaY played a role in the repair of Fe/S clusters this would imply that complex I with damaged Fe/S clusters has to be degraded in the mutant membranes.
It was proposed that CyaY is not involved in the biogenesis of Fe/S clusters but may serve as an iron chaperone binding cellular iron in order to reduce oxidative damage under oxidative stress . According to this CyaY binds redox active iron and prevents the formation of hydrogen peroxide . A deletion of CyaY leads to an increased cellular amount of reactive oxygen species which subsequently destroys Fe/S clusters in iron-sulfur proteins . If the oxidative damage of the Fe/S clusters due to the deletion of CyaY leads to the reduced content of complex I all clusters of the complex would be concerned to the same extent, as the stoichiometry of the individual Fe/S clusters is virtually identical in the preparation from wild type and the cyaY deletion strain (Fig. 4). This is rather unlikely because the accessibility of the Fe/S clusters of complex I and the polarity of their environment differ significantly . If CyaY played a role in preventing a damage of the Fe/S clusters this would again imply that complex I with damaged Fe/S clusters has to be degraded in the mutant membranes. However, it has been demonstrated that complex I with a reduced content of a single Fe/S cluster is stable in the membrane [29, 38].
Thus, it is more likely that CyaY is one component of the machinery of Fe/S cluster biogenesis as proposed [24, 41, 42]. Formation of Fe/S clusters require complex biosynthesis systems which share the involvement of cysteine desulfurases and Fe/S cluster scaffold proteins [42, 43]. The desulfurase catalyzes the reductive conversion of cysteine to alanine and sulfide. Together with iron, the sulfide is assembled to a binuclear Fe/S cluster on the scaffold protein. The dimer of the scaffold protein is able to build a tetranuclear Fe/S cluster out of the two binuclear clusters . Due to the low solubility of iron and its toxicity in the cell, the presence of an iron donor protein is most likely. The role of the protein would be to bind and hence solubilize the iron, and to transport and deliver it to the desulfurase/scaffold proteins. In this respect it should not be essential for the biosynthetic reaction. There is evidence that frataxin is the protein sought after. It was shown that frataxins are able to bind iron and exist in oligomeric forms, which seem to be the active species [20, 24, 44–47] and that they interact with the scaffold protein and the desulfurase [23, 24]. In agreement with the proposed function of the iron donor protein, it was shown that the deletion of the corresponding genes in yeast and in E. coli is not lethal [21, 25]. In accordance with these data, the deletion of cyaY leads to a reduced content of complex I and II in E. coli, most likely due to a reduced biogenesis of their Fe/S clusters, but it did not significantly change the growth rate of the mutant strain.
The E. coli frataxin homologue CyaY is located in the cytoplasm and is therefore not a structural component of complex I. A cyaY deletion mutant showed a reduced complex I and complex II production which may be a result of the disturbance of the Fe/S cluster assembly machinery. Further studies are necessary to determine if a transient interaction between the complexes of the respiratory chain and CyaY takes place during the biogenesis of the Fe/S clusters.
E. coli K-12 strains AG1, BW25113 and BW25113 cyaY::nptI were kindly provided by the Keio collection of the Nara Institute of Science and Technology (National BioResource Project (NIG, Japan): E. coli) . The plasmids pCA24NcyaY and pCA24NcyaY-gfp were obtained from the ASKA library . They contain the sequence coding for CyaY with an N-terminal 6× histidine (His)-tag (pCA24NcyaY) and an additional C-terminal GFP-fusion (pCA24NcyaY-gfp). The plasmid pGFPe nuoJ contains nuoJ engineered with a C-terminal GFP-fusion and was kindly provided by Drs. Gunnar von Heijne and Daniel Daley . The expression host BL21(DE3) for pGFPe nuoJ was purchased from Novagen. Strains were grown in LB medium at 37°C until early stationary phase. Cells used for fluorescence microscopy were grown in M9 minimal medium with 30 mM mannitol as carbon source at 25°C. Chloramphenicol (100 μg/mL) and kanamycin (50 μg/ml) were supplemented when necessary. Expression of pBADcyaY-gfp (see below) was induced by adding 0.2% (w/v) L-arabinose to the media and repressed in the presence of 0.2% (w/v) D-glucose. Expression of pCA24NcyaY and pGFPe nuoJ was induced by an addition of 1 mM isopropyl-β-D-thiogalactopyranoside.
The cyaY-gfp fusion was PCR-amplified from pCA24NcyaY-gfp using Phusion DNA Polymerase (Finnzymes) and primers XbaI-cyaY (5'AGTTCTAGAAGG AGGAATTCACCATGAACACAGTGAATTTCATCGCCTG) and gfp-HindIII (5'-AGTAAGCTTGCAGGTCGACCCTTAGCG) and cloned in pBAD33  using the same primers. The forward primer contains a synthetic ribosomal binding site AGGAGG 8 nt upstream of the initiation codon. The PCR product was cut with XbaI and HindIII and ligated to XbaI-HindIII sites of pBAD33 downstream of the araBAD promotor.
Complex I was isolated similarly to the procedure described . All steps were carried out at 4°C. 76 g cells were resuspended in a 5-fold volume of 50 mM MES/NaOH, 0.1 mM phenylmethanesulfonyl fluoride, pH 6.0, with 10 μg/mL DNAseI and 50 μg/mL lysozyme and disrupted by a single pass through a French Pressure cell (SLM Aminco) at 110 MPa. Cell debris was removed by centrifugation at 36000 × g for 20 min and cytoplasmic membranes were obtained by centrifugation at 250000 × g for 1 h. The membranes were resuspended in 50 mM MES/NaOH, 50 mM NaCl, pH 6.0 at a protein concentration of 80 mg/mL. n-Dodecyl-β-D-maltopyranoside (DDM, AppliChem) was added to a final concentration of 3% and the solution was gently homogenized and centrifuged for 20 min at 250000 × g. The supernatant was applied to a 120 mL Fractogel EMD TMAE Hicap M (Merck) column equilibrated in 50 mM MES/NaOH, 50 mM NaCl and 0.1% DDM, pH 6.0. The column was eluted with a 150 mL linear gradient of 150–350 mM NaCl in 50 mM MES/NaOH, 0.1% DDM, pH 6.0 at a flow rate of 15 mL/min. Fractions containing NADH/ferricyanide oxidoreductase activity were combined, concentrated by precipitation with 9% (w/v; final concentration) poly(ethylene glycol) 4000 and dissolved in 5 mL 50 mM MES/NaOH, 50 mM NaCl and 0.1% DDM, pH 6.0. The proteins were loaded onto a 80 mL Source 15Q (GE Healthcare) column equilibrated in 50 mM MES/NaOH, 50 mM NaCl and 0.1% DDM, pH 6.0. The column was eluted with a 500 mL linear gradient of 125–275 mM NaCl in 50 mM MES/NaOH, 0.1% DDM, pH 6.0 at a flow rate of 5 mL/min. Fractions containing NADH/ferricyanide oxidoreductase activity were pooled and concentrated with a 100 kDa MWCO Amicon Ultra-15 centrifugal filter (Millipore). The concentrated protein solution was subjected to size-exclusion chromatography on a 450 mL Sephacryl S-300 HR (GE Healthcare) column in 50 mM MES/NaOH, 50 mM NaCl, and 0.1% DDM, pH 6.0, at a flow rate of 20 mL/h. Peak fractions of NADH/ferricyanide oxidoreductase activity were combined and stored at -80°C.
His-tagged CyaY was isolated from strain AG1/pCA24NcyaY. For SDS-PAGE analysis 0.24 g wet cells were treated with Bugbuster protein extraction reagent (Novagen) according to manufacturer's recommendations. Cell debris and lipids were removed by centrifugation at 150000 × g for 15 min at 4°C and the supernatant was adjusted to 20 mM imidazole and loaded onto a 100 μL His Spin-Trap column (GE Healthcare). After washing with 600 μL 20 mM imidazole in 20 mM Na3PO4/HCl pH 7.4 and 600 μL 500 mM imidazole in 20 mM Na3PO4/HCl pH 7.4 the protein was eluted with 600 μL 1 M imidazole in 20 mM Na3PO4/HCl pH 7.4.
For determination of iron binding capacity His-tagged CyaY was isolated from 2 g wet cells. All steps were carried out at 4°C. The cells were resuspended in a 10-fold volume of 20 mM Na3PO4/HCl, 20 mM imidazole, 500 mM NaCl 0.1 mM phenylmethanesulfonyl fluoride, pH 7.4 (binding buffer), with 10 μg/mL DNAseI and 50 μg/mL lysozyme and disrupted by a single pass through a French Pressure cell (SLM Aminco) at 110 MPa. Cell debris and membranes were removed by centrifugation at 250000 × g for 1 h. The supernatant was loaded onto a 15 mL ProBond Ni2+-IDA column equilibrated in binding buffer. The column was washed with 80 mL of binding buffer and proteins were eluted with a 60 mL linear gradient of 20–1000 mM imidazole in binding buffer. Fractions were analyzed by SDS-PAGE and those containing His-tagged CyaY were combined and concentrated by a 3 kDa MWCO Vivaspin 20 centrifugal filter (Vivascience).
EPR spectroscopy was performed with a Bruker EMX 1/6 spectrometer operating at X-band (9.2 GHz) according to . The magnetic field was calibrated using a strong or a weak pitch standard. The isolated complex I (3 mg/mL) was reduced with a few grains of dithionite in the presence of a 1000-fold molar excess of NADH.
Complex I activity in the cytoplasmic membranes was measured either as NADH/ferricyanide oxidoreductase activity or as d-NADH oxidase activity as described [27, 29]. The d-NADH oxidase activity was inhibited by addition of 20 μM annonin VI, a specific complex I inhibitor . The NADH:decyl-ubiquinone oxidoreductase activity of isolated complex I was determined as described . The succinate oxidase activity of cytoplasmic membranes was measured with a clark oxygen electrode. To remove tightly bound oxaloacetate at the active site and to express the full catalytic activity of succinate dehydrogenase, cytoplasmic membranes corresponding to 600 μg protein were incubated for 10 min in 2 mL 30 mM Na2HPO4/HCl pH 7.4 at 30°C . The reaction was initiated by the addition of 20 mM succinate and inhibited in the presence of 40 mM malonate.
The binding of iron to the overproduced and His-tagged CyaY was determined according to . Purified CyaY was incubated with a 15-fold molar excess of Fe3+ in form of ferric ammonium citrate under aerobic conditions at 4°C for 1 hour. The mixtures were centrifuged at 16000 × g and desalted on a 20 mL Sephadex G25 superfine column (Amersham Pharmacia) equilibrated in 50 mM Tris/HCl, 50 mM NaCl, pH 7.0. The iron content of the preparation before and after loading with Fe3+ was determined according to .
Cells at mid-exponential phase were placed on a microscope slide covered with a pad of 1% (w/v) agarose in M9-mannitol minimal media. A cover slip was placed on the cells and images were acquired with a Axio Imager A.1 fluorescence microscope (Zeiss) at 1000× magnification. Pictures were acquired with a digital CCD camera and processed with Metamorph 4.6 (Universal Imagin Corp., USA). The cytoplasmic membranes were stained with 1 nM N-(3-triethylammoniumpropyl)-4-(p-diethylaminophenylhexatrienyl)pyridinium dibromide (FM4-64).
Protein concentration was measured either by the biuret or the Bradford method  using BSA as standard. SDS-PAGE was performed according to the protocol of Schägger and von Jagow , using a 10% T, 3% C separating gel. The concentration of the isolated proteins was determined by the absorbance at 280 nm using an extinction coefficient of 764 mM-1cm-1 for complex I and 30 mM-1cm-1 for CyaY derived from their DNA sequence. Sucrose-gradient centrifugation in the presence of 0.1% DDM was performed as described . Proteins separated by SDS-PAGE were electroblotted onto 0.45 μm pore size nitrocellulose membrane (Schleicher und Schüll) for western blot analysis . An Anti-Penta-His antibody (Qiagen) was used at a 1:400 dilution for detection. The membrane was incubated for 1 h at 25°C with the primary antibody.
electron paramagnetic resonance
This work was supported by the Deutsche Forschungsgemeinschaft. We are grateful to the people working at the Keio Collection and ASKA Library in Japan and Dr. Daniel Daley for the provision of strains and plasmids. We thank Linda Böhm for critical reading of the manuscript.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.